Sections and Successions in Successful Songs: A Prototype Approach to Form in Rock Music by Trevor Owen de Clercq Submitted in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy Supervised by Professor David Temperley Department of Music Theory Eastman School of Music University of Rochester Rochester, NY 2012
317
Embed
Sections and Successions in Successful Songs: A Prototype ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Sections and Successions in Successful Songs:
A Prototype Approach to Form in Rock Music
by
Trevor Owen de Clercq
Submitted in Partial Fulfillment
of the
Requirements for the Degree
Doctor of Philosophy
Supervised by
Professor David Temperley
Department of Music Theory
Eastman School of Music
University of Rochester
Rochester, NY
2012
ii
Curriculum Vitae
Trevor Owen de Clercq was born in Montreal, Canada on March 2, 1975. He grew up
in Naples, FL, and graduated salutatorian from Naples High School in 1992. He then
attended Cornell University (1992-1996), where he graduated cum laude with a Bachelor’s
degree in Music. After college, he worked as a Grants Manager for the Harvard Medical
School (1996-1998). He then enrolled in a Tonmeister program at New York University
(1998), from which he earned a Master’s degree in Music Technology (2000). Afterwards, he
worked in various New York City recording studios, most notably Right Track Recording
(2001-2002), where he supported albums by artists such as Mariah Carey, Fabolous, James
Taylor, Britney Spears, Nas, Pat Metheny, and Mark O’Connor. He then worked as a
technical support specialist at The New School (2002-2006). During this time, he also
earned (via distance learning) an Associate’s degree in Electrical Engineering Technology
from the Cleveland Institute of Electronics (2004). In 2006, he entered the Music Theory
program at the Eastman School of Music, from which he received a Master’s degree in 2008.
The recipient of a Sproull Fellowship, he has served as a graduate instructor in both the
Department of Music Theory (2008-2010) and the Department of Electrical and Computer
Engineering at the University of Rochester (2009). He has pursued his research under the
direction of Professor David Temperley. His most recent position is as Adjunct Assistant
Professor at Hofstra University (Hempstead, NY), where he teaches written theory and
musicianship coursework to undergraduates.
List of Publications:
de Clercq, Trevor and David Temperley. 2011. “A corpus analysis of rock harmony.” Popular
Music 30 (1): 47-70.
iii
Acknowledgments
The final form of this dissertation is due in large part to the incredibly precise and
clear insights of its primary advisor, David Temperley. Thank you, Davy, for all your hard
work on this project. The other members of my committee, Seth Monahan and Chris Doll,
also have my sincerest gratitude for their assistance. Thanks are due as well to John Covach,
who provided helpful navigational advice during my time at sea in the early stages of this
project. Overall, I would like to thank the entire Eastman music theory community for
creating a highly supportive academic environment. Robert L. and Mary L. Sproull provided
generous financial aid during my five years in residence. Dave Headlam helped as well by
hiring me as his laboratory assistant and connecting me with various employment
opportunities in the ECE Department. On a personal level, I would like to thank my family
for enduring my long and sinuous path in the pursuit of knowledge. My mom, Suzanne, and
dad, Ted, have never failed to provide good advice when I needed it. I will also be forever
indebted to my honorary Godmother, the late Clarice Holtz. Finally, to Sarah and little Rufus:
The world would be a much colder and lonelier place without you. Thank you both for the
love, compassion, and sympathy through the years.
iv
Abstract
This dissertation tackles the open question of how listeners perceive form in rock
music. Extant theories of form offer only limited explanations as to how we categorize the
various components of a rock song and how these choices affect our overall understanding of
form in this repertoire. Research in the field of cognitive psychology shows that our
categorization process involves prototype effects. Consequently, this dissertation employs a
prototype-based approach to form in rock music. A central task in this regard is the
development of a broader understanding of the prototypical instantiations of section roles,
including verse, chorus, refrain, bridge, solo, prechorus, intro, outro, and link. Using this
information, we can see how these section roles interact with different organizational
schemes. Three main organizational schemes are used as points of reference: the 12-bar
blues, the 16-bar SRDC, and the 32-bar AABA. As various attributes of these schemes change
from song to song, we see conversions of these schemes from one section role to another. As
a result, we expose possible paths in the historical development of form within rock music.
During this conversion process, various types of ambiguities between section roles can be
found. Often, in fact, the choice between one section role and another inherently represents a
false dilemma. The notion of blends – which describe amalgamations of two or more section
roles – becomes useful to track these ambiguous cases. Ultimately, this research appraises
those factors that drive analytical practice and attunes us to the complex ways that real-world
songs engage with our expectations and sustain our fascination.
v
Table of Contents
List of Examples vi vi Chapter 1: Prologue 1 1 Chapter 2: Background 10 2.1: Introduction 10 2.2: Cognitive Science and Categories 10 2.3: Prototypes and Music Theory 17 2.4: Music Theory and Form 20 2.5: Music Cognition and Form 23 2.6: Form, Music Theory, and Rock Music 24 2.7: Summary 33 Chapter 3: Roles 34 3.1: Introduction 34 3.2: Verse and Chorus 38 3.3: Refrain 57 3.4: Bridge and Solo 70 3.5: Prechorus 89 3.6: Intro, Outro, and Link 99 3.7: Summary 110 Chapter 4: Conversions 117 4.1: Introduction 117 4.2: The Blues 123 4.3: SRDC 153 4.4: AABA 178 4.5: Summary 211 Chapter 5: Blends 213 5.1: Introduction 213 5.2: Verse Blends (Part 1) 215 5.3: Bridge Blends 221 5.4: Chorus Blends 238 5.5: Verse Blends (Part 2) 268 X 5.6: Summary 284 Chapter 6: Epilogue 287 Bibliography 294 Musical Sources 300
vi
List of Examples
3.2.01: “Just the Way You Are” (Bruno Mars, 2010); verse 44 3.2.02: “Just the Way You Are” (Bruno Mars, 2010); chorus 44 3.2.03: “Little Red Corvette” (Prince, 1982); verse 48 3.2.04: “Little Red Corvette” (Prince, 1982); chorus 49 3.2.05: “Rockin’ in the Free World” (Neil Young, 1989); verse 51 3.2.06: “Rockin’ in the Free World” (Neil Young, 1989); chorus 51 3.2.07: “Just What I Needed” (The Cars, 1978); verse 54 3.2.08: “Just What I Needed” (The Cars, 1978); chorus 54 3.3.01: “Stand by Me” (Ben E. King, 1961); verse material with refrain 60 3.3.02: “Stand by Me” (Ben E. King, 1961); form chart 61 3.3.03: Phrase structure of a prototypical tail refrain 61 3.3.04: “All I Have to Do Is Dream” (The Everly Brothers, 1958); verse 63 3.3.05: “I’ve Got a Tiger by the Tail” (Buck Owens, 1965); chorus 64 3.3.06: “Old Time Rock and Roll” (Bob Seger & The Silver Bullet Band, 1978); chorus 65 3.3.07: “Old Time Rock and Roll” (Bob Seger & The Silver Bullet Band, 1978); verse 67 3.3.08: “Every Breath You Take” (The Police, 1983); opening material 68 3.3.09: “Every Breath You Take” (The Police, 1983); closing material 69 3.4.01: “Handy Man” (Jimmy Jones, 1960); form chart 74 3.4.02: “Handy Man” (Jimmy Jones, 1960); A section 75 3.4.03: “Handy Man” (Jimmy Jones, 1960); B section 75 3.4.04: Generic phrase organization for a classic bridge 76 3.4.05: Prototypical harmonic realizations for a classic bridge 77 3.4.06: 32 songs with classic bridge sections 77 3.4.07: “Ticket to Ride” (The Beatles, 1965); bridge (B section) 79 3.4.08: “You Can’t Do That” (The Beatles, 1964); bridge (B section) 80
List of Examples vii
3.4.09: “1979” (The Smashing Pumpkins, 1995); bridge 82 3.4.10: “1979” (The Smashing Pumpkins, 1995); core AABA pattern 82 3.4.11: “More Than a Feeling” (Boston, 1976); form chart in Covach 2005 (74) 84 3.4.12: “Whole Lotta Love” (Led Zeppelin, 1969); form chart in Covach 2003 (183) 85 3.4.13: “Every Breath You Take” (The Police, 1983); classic bridge 86 3.4.14: “Every Breath You Take” (The Police, 1983); form chart in Covach 2005 (75) 86 3.4.15: “Every Breath You Take” (The Police, 1983); modern bridge 87 3.4.16: “Every Breath You Take” (The Police, 1983); alternative grouping 88 3.5.01: “You Belong with Me” (Taylor Swift, 2008); verse 93 3.5.02: “You Belong with Me” (Taylor Swift, 2008); chorus 93 3.5.03: “You Belong with Me” (Taylor Swift, 2008); prechorus 94 3.5.04: “You Belong with Me” (Taylor Swift, 2008); form chart 95 3.5.05: “Building a Mystery” (Sarah McLachlan, 1997); verse 97 3.5.06: “Building a Mystery” (Sarah McLachlan, 1997); chorus 97 3.5.07: “Building a Mystery” (Sarah McLachlan, 1997); prechorus 97 3.6.01: Four different treatments of an instrumental interlude (link) a) Interlude as post-chorus: “Thank You (Falettinme Be Mice Elf Agin)” (Sly and The Family Stone, 1969); form chart in Covach 2009 (370-1) 102 b) Interlude as pre-verse: “One” (Metallica, 1988); form chart in Covach 2009 (494-5) 103 c) Interlude as verse: “More Than a Feeling” (Boston, 1976); form chart in Covach 2009 (418-9) 103 d) Interlude as chorus: “All I Wanna Do” (Sheryl Crow, 1994); form chart in Covach 2009 (530-1) 103 3.6.02: “In Bloom” (Nirvana, 1991); form chart 104 3.6.03: “In Bloom” (Nirvana, 1991); alternative form chart 105 3.6.04: “In Bloom” (Nirvana, 1991); chorus into link 106 3.6.05: “My Happy Ending” (Avril Lavigne, 2004); form chart 108 3.6.06: “My Happy Ending” (Avril Lavigne, 2004); chorus into link 109 3.7.01: “You Might Think” (The Cars, 1984); form chart 112 4.1.01: Phrase rhythms in rock music, à la Stephenson 2002 120
List of Examples viii
4.1.02: Melodic phrase relationships 121 4.2.01: Harmonic realizations of the 12-bar blues in various authors 124 4.2.02: “Crossroads” (Cream, 1968); opening vocal material 125 4.2.03: Classic 12-bar blues phrase structure 125 4.2.04: “Shake, Rattle and Roll” (Big Joe Turner, 1954); verse 127 4.2.05: “Shake, Rattle and Roll” (Big Joe Turner, 1954); chorus 127 4.2.06: Phrase structure for chorus of “Shake, Rattle and Roll” 128 4.2.07: “Ko Ko Mo (I Love You So)” (The Crew Cuts, 1955); verse 129 4.2.08: “Ko Ko Mo (I Love You So)” (The Crew Cuts, 1955); chorus 129 4.2.09: Phrase structure for chorus of “Ko Ko Mo (I Love You So)” 129 4.2.10: “Maybellene” (Chuck Berry, 1955); chorus 130 4.2.11: “Maybellene” (Chuck Berry, 1955); verse 131 4.2.12: “Johnny B. Goode” (Chuck Berry, 1958); verse 132 4.2.13: Phrase structure for verse of “Johnny B. Goode” 132 4.2.14: “Johnny B. Goode” (Chuck Berry, 1958); chorus 133 4.2.15: “Hound Dog” (Elvis Presley, 1956); form chart with lyrics 133 4.2.16: “Hound Dog” (Elvis Presley, 1956); section 1 134 4.2.17: “Hound Dog” (Elvis Presley, 1956); section 2 incipit 134 4.2.18: “Evil” (Howlin’ Wolf, 1954); opening vocal material 136 4.2.19: Hybrid 12-bar blues phrase structure 137 4.2.20: “Strange Brew” (Cream, 1967); opening 12-bar blues 138 4.2.21: “Long Tall Sally” (Little Richard, 1956); opening vocal material 139 4.2.22: “Ooby Dooby” (Roy Orbison, 1956); opening vocal material 140 4.2.23: “Boys” (Shirelles, 1960); opening vocal material 141 4.2.24: “Boys” (Shirelles, 1960); chorus 141 4.2.25: “Boys” (Shirelles, 1960); form chart 142 4.2.26: “Blue Suede Shoes” (Elvis Presley, 1956); opening vocal material 143
List of Examples ix
4.2.27: “Blue Suede Shoes” (Elvis Presley, 1956); 16-bar blues opening 143 4.2.28: Hybrid 16-bar blues phrase structure 144 4.2.29: “Jailhouse Rock” (Elvis Presley, 1957); main section(s) 145 4.2.30: “Taxman” (The Beatles, 1966); opening vocal material 147 4.2.31: “Taxman” (The Beatles, 1966); middle vocal material 149 4.2.32: “Taxman” (The Beatles, 1966); closing vocal material 150 4.2.33: “Taxman” (The Beatles, 1966); form chart 151 4.2.34: Relationships between blues schemes and section roles 152 4.3.01: SRDC pattern within a hybrid 16-bar blues phrase structure 154 4.3.02: 8-bar SRDC pattern in “From Me to You” (The Beatles, 1963) 156 4.3.03: “I’ll Cry Instead” (The Beatles, 1964); SRDC in main material 157 4.3.04: Classic 16-bar SRDC phrase structure 157 4.3.05: “Please Please Me” (The Beatles, 1963); SRDC pattern 158 4.3.06: “Dream Lover” (Bobby Darin, 1959); SRDC in main material 160 4.3.07: “Ticket to Ride” (The Beatles, 1965); SRDC in main section(s) 161 4.3.08: “Ticket To Ride” (The Beatles; 1965); form chart 162 4.3.09: “Drive My Car” (The Beatles, 1965); SRDC in main section(s) 163 4.3.10: SRDC phrase structure in “Drive My Car” (The Beatles, 1965) 164 4.3.11: “Drive My Car” (The Beatles, 1965); form chart 165 4.3.12: “La-La (Means I Love You)” (The Delfonics, 1968); main unit 166 4.3.13: Abstraction of SRDCC structure in “La-La (Means I Love You)” 166 4.3.14: “I Heard It Through the Grapevine” (Marvin Gaye, 1968); main repeating unit 168 4.3.15: Vocal phrase organization in “I Heard It Through the Grapevine” 168 4.3.16: “I Can See for Miles” (The Who, 1967); main material 169 4.3.17: “I Can See for Miles” (The Who, 1967); extended conclusion 170 4.3.18: “I Can See for Miles” (The Who, 1967); potential tail refrain 170 4.3.19: Metric reinterpretation 171
List of Examples x
4.3.20: Hypermetric reinterpretation 172 4.3.21: “(I Can’t Get No) Satisfaction” (The Rolling Stones, 1965); main vocal material 174 4.3.22: “You’ve Lost That Lovin’ Feelin’” (The Righteous Brothers, 1964); opening material 175 4.3.23: “You’ve Lost That Lovin’ Feelin’” (The Righteous Brothers, 1964); chorus 176 4.3.24: Relationships between SRDC schemes and section roles 177 4.4.01: “Will You Love Me Tomorrow” (The Shirelles, 1960); A section 181 4.4.02: “Will You Love Me Tomorrow” (The Shirelles, 1960); B section 181 4.4.03: “Will You Love Me Tomorrow” (The Shirelles, 1960); form chart 182 4.4.04: “Love Me Do” (The Beatles, 1963); form chart 184 4.4.05: “Love Me Do” (The Beatles, 1963); A section 185 4.4.06: “I’m Walkin’” (Fats Domino, 1957); form chart 186 4.4.07: “I’m Walkin’” (Fats Domino, 1957); B section 186 4.4.08: “I’m Walkin’” (Fats Domino, 1957); A section 186 4.4.09: Phrase organization for a classic 8-bar A section 187 4.4.10: “True Love Ways” (Buddy Holly, 1960); first two A sections 188 4.4.11: “True Love Ways” (Buddy Holly, 1960); organization of consequent A section 189 4.4.12: “True Love Ways” (Buddy Holly, 1960); B and A sections 190 4.4.13: “True Love Ways” (Buddy Holly, 1960); form chart 190 4.4.14: “Can’t Buy Me Love” (The Beatles, 1964); A section 191 4.4.15: “Can’t Buy Me Love” (The Beatles, 1964); form chart 192 4.4.16: “Can’t Buy Me Love” (The Beatles, 1964); B section 193 4.4.17: “Blitzkrieg Bop” (Ramones, 1976); A section 194 4.4.18: “Blitzkrieg Bop” (Ramones, 1976); B section 195 4.4.19: “Blitzkrieg Bop” (Ramones, 1976); form chart 195 4.4.20: “Suspicious Minds” (Elvis Presley, 1969); verse material 197 4.4.21: “Suspicious Minds” (Elvis Presley, 1969); chorus material 198
List of Examples xi
4.4.22: “Suspicious Minds” (Elvis Presley, 1969); form chart 199 4.4.23: “Sin City” (The Flying Burrito Brothers, 1969); verse material 201 4.4.24: “Sin City” (The Flying Burrito Brothers, 1969); chorus material 201 4.4.25: “Handy Man” (Jimmy Jones, 1960); A section 203 4.4.26: “Handy Man” (Jimmy Jones, 1960); B and A sections 204 4.4.27: “Handy Man” (Jimmy Jones, 1960); form chart 204 4.4.28: “Handy Man” (James Taylor, 1977); A section 205 4.4.29: “Handy Man” (James Taylor, 1977); B and A sections 205 4.4.30: “Handy Man” (James Taylor, 1977); form chart 206 4.4.31: “God Save the Queen” (The Sex Pistols, 1977); form chart 207 4.4.32: “God Save the Queen” (The Sex Pistols, 1977); A1 section 208 4.4.33: “God Save the Queen” (The Sex Pistols, 1977); A3 section 208 4.4.34: Relationships between AABA-derived organizational schemes and section roles 210 5.2.01: “Smells Like Teen Spirit” (Nirvana, 1991); verse material 216 5.2.02: “Smells Like Teen Spirit” (Nirvana, 1991); chorus material 216 5.2.03: “Smells Like Teen Spirit” (Nirvana, 1991); prechorus candidate 217 5.2.04: “Run to You” (Bryan Adams, 1984); main material 219 5.3.01: “Smoke on the Water” (Deep Purple, 1972); form chart in Covach 2005 (73) 222 5.3.02: “Angel” (Aerosmith, 1987); form chart 223 5.3.03: “Buddy Holly” (Weezer, 1994); form chart 224 5.3.04: “Smells Like Teen Spirit” (Nirvana, 1991); form chart in Covach 2009 (512) 225 5.3.06: “Wrapped Around Your Finger” (The Police, 1983); verse 226 5.3.07: “Wrapped Around Your Finger” (The Police, 1983); form chart 227 5.3.08: “Wrapped Around Your Finger” (The Police, 1983); verse/bridge 228 5.3.10: “Him” (Rupert Holmes, 1980); form chart in Stephan-Robinson 2009 (75-6) 231 5.3.11: “Him” (Rupert Holmes, 1980); prechorus candidate 232
List of Examples xii
5.3.12: Form in three songs from Boys Like Girls (Boys Like Girls, 2006) 235 5.4.01: “Sympathy for the Devil” (The Rolling Stones, 1968); second iteration of main unit 239 5.4.02: “Sympathy for the Devil” (The Rolling Stones, 1968); phrase rhythm 240 5.4.03: “We Will Rock You” (Queen, 1977); verse 242 5.4.04: “We Will Rock You” (Queen, 1977); refrain/chorus 242 5.4.05: “1999” (Prince, 1982); verse material 243 5.4.06: “1999” (Prince, 1982); refrain/chorus 243 5.4.07: “1999” (Prince, 1982); form chart 244 5.4.08: “Shop Around” (The Miracles, 1960); form chart 245 5.4.09: “Shop Around” (The Miracles, 1960); 8-bar post-B A material 245 5.4.10: “Shop Around” (The Miracles, 1960); opening A material 246 5.4.11: “Shop Around” (The Miracles, 1960); alternative metric organization for refrain 247 5.4.12: “Train in Vain (Stand by Me)” (The Clash, 1979); main material 248 5.4.13: “Take Me to the River” (Al Green, 1974); form chart in Temperley 2010 250 5.4.14: “Jump” (Van Halen, 1984); verse 252 5.4.15: “Jump” (Van Halen, 1984); prechorus (plus overlap) 253 5.4.16: “Jump” (Van Halen, 1984); link/chorus 253 5.4.17: “Jump” (Van Halen, 1984); form chart 254 5.4.18: “Summer of ‘69” (Bryan Adams, 1984); verse 256 5.4.19: “Summer of ‘69” (Bryan Adams, 1984); second part (part 2) 256 5.4.20: “Summer of ‘69” (Bryan Adams, 1984); form chart 257 5.4.21: “Summer of ‘69” (Bryan Adams, 1984); link or chorus? 257 5.4.22: “Born to Run” (Bruce Springsteen, 1975); verse 260 5.4.23: “Born to Run” (Bruce Springsteen, 1975); second section 261 5.4.24: “Communication Breakdown” (Led Zeppelin, 1969); verse 263 5.4.25: “Communication Breakdown” (Led Zeppelin, 1969); chorus 263
List of Examples xiii
5.4.26: “Communication Breakdown” (Led Zeppelin, 1969); form chart 264 5.4.27: “Screen Door” (Uncle Tupelo, 1990); form chart in Covach 2009 (564) 265 5.4.28: “Screen Door” (Uncle Tupelo, 1990); verse 266 5.4.29: “Screen Door” (Uncle Tupelo, 1990); part 2 267 5.5.01: “Tears in Heaven” (Eric Clapton, 1992); form chart in Temperley 2011 269 5.5.02: “Tears in Heaven” (Eric Clapton, 1992); Temperley’s verse 270 5.5.03: “Tears in Heaven” (Eric Clapton, 1992); Temperley’s chorus 270 5.5.04: “That Feel” (Tom Waits, 1992); first section (A) 273 5.5.05: “That Feel” (Tom Waits, 1992); second section (B) 273 5.5.06: “That’ll Be the Day” (The Crickets, 1957); form chart in Covach 2009 (104) 275 5.5.07: “That’ll Be the Day” (The Crickets, 1957); Covach’s chorus 276 5.5.08: “That’ll Be the Day” (The Crickets, 1957); Covach’s verse 276 5.5.09: “Hey Good Lookin’” (Hank Williams, 1951); bridge 277 5.5.10: “Today” (The Smashing Pumpkins, 1993); form chart 279 5.5.11: “Today” (The Smashing Pumpkins, 1993); initial A material (A1) 279 5.5.12: “Today” (The Smashing Pumpkins, 1993); initial of B material (B1) 280 5.5.13: “Today” (The Smashing Pumpkins, 1993); second B material (B2) 280 5.5.14: “Today” (The Smashing Pumpkins, 1993); third A material (A3) 282 5.5.15: “Today” (The Smashing Pumpkins, 1993); third B material (B3) 283 5.6.01: Common blends involving neighboring section roles 285
1
Chapter 1: Prologue
Form has been the subject of intensive and diverse theorization throughout the
history of music. More so than any other aspect of music, form allows for a wide variety of
approaches because theories of form inherently engage with every possible musical parameter
– harmony, counterpoint, motive, rhythm, meter, etc. – and thereby inherit the variety of
approaches associated with these domains. The numerous methodologies that scholars have
proposed to account for the form(s) found in the first movements of classical sonatas testify
to the difficulties posed by just a single form category (see Burnham 2002). Spanning more
than two centuries, theories of phrase and rhetoric (Mattheson and Koch), theme and
development (Reicha and Marx), formal functions (Riemann, Ratz, and Caplin), and the “Will
of the Tones” (Kurth and Schenker) offer a panoply of answers to the basic question: “What is
form?”
In contrast to common-practice music, form in rock music may seem relatively
straightforward. Indeed, theorists sometimes express basic satisfaction with the modern
form terminology of rock and its analytical usage. For example, Jocelyn Neal writes: “The
mere act of labeling sections of a song is little more than a rote exercise, one that is easily and
frequently taught to undergraduate students of popular music” (2007, 44). With a similar
connotation, Allan Moore writes that section labels such as verse, refrain, chorus, and
bridge are “categories frequently used by writers and performers, and their ubiquity is
sufficient to ensure their analytical value” (2001, 52). Yet despite the analytical value that
Moore assumes these categories hold, he devotes very little ink to explaining what these labels
mean or how they should be applied. The reader is left with only the brief glossary entries for
these terms provided at the end of his book. Presumably, the meanings of these terms are
self-evident.
As we examine published analyses for a variety of rock songs, however, it quickly
becomes clear that the use of section labels is much less straightforward than these authors
suggest. In many cases, different theorists provide different (if not contradictory) analyses of
the same song (examples can be found throughout this dissertation). The extent and nature
of this disagreement varies. In some cases, section labels differ; in other cases, section labels
coincide, but their boundaries do not; often the issue is simply whether a particular group of
measures deserves to be a separate section at all. Some scholars might not be concerned with
this disagreement, as one could say that differing interpretations of a musical work testify to
the richness and complexity of the work itself; undoubtedly, much great music supports
multiple readings. But the large number of conflicting interpretations gives us cause to
reconsider the presupposed self-evident nature of our section labels.
Chapter 1: Prologue
2
Recently, in fact, a number of theorists have bemoaned the state of terminology
within the field. For example, Christopher Endrinal writes that “[in examining rock music,]
one of the difficulties has been the terminology used to describe the form of rock songs and
the fact that… little has been written regarding some general definitions of form in rock
music” (2008, 61). Similarly, Paul Harris writes that “there are no settled definitions for what
constitutes, and thus distinguishes, verses, choruses, and bridges, and the infinitely variable
interludes acting as spacers between the main structural components” (2006, 62).
Underlying these complaints is the general notion that, without a shared set of definitions for
these terms, we cannot effectively communicate to one another our ideas about form in rock
music. In other words, our understanding of form in rock relies heavily on these section
labels and the definitions we assign them.
The importance of providing definitions for section labels may thus appear to be a
central task for theorists of rock music. In this regard, it is worth considering what
constitutes a definition. In formal terms, there are two main ways that a definition can be
constructed. These are the intensional and extensional techniques (Hurley 2008, 97ff). With
the extensional technique, meaning is assigned to a term by indicating which items belong to
the category it encompasses. In essence, this technique involves simply a list of members
(whether individuals or groups). For example, the statement “The Baltic states are Estonia,
Latvia, and Lithuania” defines the term “Baltic states” through a complete inventory of its
members. In contrast, a term defined via the intensional technique lays out the essential
properties of the term itself. To put it another way, the intensional technique provides the
means necessary to identify its members. For example, the statement “Ice is frozen water”
states exactly what qualities are sufficient to classify something as “ice.” If something is
frozen water, then it is ice. Conversely, if something is not frozen water, then it is not ice. It
should be noted that with either the intensional or extensional technique, there is a strict
process of inclusion and exclusion. Something either is or is not a “Baltic state” or “ice,” for
instance.
Reading through the work of music theorists, we often find evidence of a definition-
based approach, usually using the intensional technique. Take, for example, Endrinal’s
analysis of “Elevation” (U2, 2000). This song contains a middle passage (beginning at 2:11)
for which one might consider the label “bridge” to be appropriate. But Endrinal explicitly
denies this reading. His reasoning is clear, as he writes, “[T]his section does not have
transitional function; therefore, it cannot be called a ‘bridge’” (2008, 78; emphasis in the
original). To Endrinal, therefore, bridge quality relies on this essential element of transition,
without which we must abandon the term. Bradley Osborn takes a similar tack in his analysis
of the song “Don’t Stop Believin’” (Journey, 1981). He states that, although most people
Chapter 1: Prologue
3
would consider the title-containing final passage (beginning around 3:21) to be the chorus of
the song, this passage “is not a chorus” (2010, 19 and 111-2; emphasis in the original). Here
again, we find an author denying a section label to a span of music despite apparent evidence
to the contrary. In particular, Osborn explicitly divorces our perception of a passage from
what that passage is (or is not). Note that both authors work under the assumption that
category membership is an either/or scenario. In so doing, they strongly imply that certain
elements are essential in the membership assignment process.
In many cases, though, we find that theorists ascribe non-essential properties to
section types. For instance, Everett states that a bridge is a section “often beginning in an
area other than tonic and usually leading to a dominant transition” (2001, 363). In a similar
manner, he writes that a chorus is a section type “nearly always affirming the tonic [and]
usually appearing in the song’s interior” (364). Although Mark Spicer refers to these
statements by Everett as “precise definitions” (2005, ¶10), it seems that Everett’s statements
are not very definition-like at all. With the semantic hedges of “usually,” “often,” and “nearly
always,” Everett implies that some particular chorus, for example, might not affirm tonic or
might not appear in a song’s interior. These qualities (harmony and position) thus do not
appear to be essential elements of a chorus. The implication is rather that these elements are
important to our understanding of chorus sections but are not determining factors. Instead
of offering definitions, therefore, Everett is providing information as to the usual
configurations of various section types in rock music. Why, one might wonder, does Everett
not define these section labels more strictly? As we will see, the reasons for this apparent
looseness derive from basic factors of human cognition.
A prototype approach
Since the 1970s, cognitive scientists have become aware that definitions do not model
very well the way we understand the words and labels we use. The process by which we label
something, in fact, turns out to be extremely complex – more complex than can be accounted
for by any definition. Seminal work in this regard was done by Eleanor Rosch (e.g., 1973) and
her colleagues. This research showed that when we categorize something (whether it be a
physical object, a temporal event, or a social relationship), we rely less on definitions of these
categories and more on abstract comparisons with what are judged to be the best examples of
that category. (This and related research will be discussed in more detail in the following
chapter.) Rosch referred to such abstract comparisons as evidence of prototype effects, since
certain category members are rated as more representative (i.e., prototypical) of the category
than others. Subsequent psychological research has affirmed these findings. As a result,
Chapter 1: Prologue
4
much modern work within the field of cognitive science has taken a prototype-based
approach to conceptual categorization.
A prototype-based approach is founded on the notion that there are no essential
elements for our conception of a category. Rather, we understand categories through a
panoply of attributes, none of which are required for membership in that category.
Wittgenstein’s notion of family resemblance relationships (1953) is instructive here. Objects
having attributes AB, BC, and CD may – because of these attributes – fall into the same
category, yet some of the objects in this same category (those with attributes AB and CD) have
no overlapping qualities. No essential element for that category can thus be said to exist.
Certain attributes, of course, may be more perceptually central than others. Yet peripheral
attributes can also trigger our perception of a category, and an amalgamation of peripheral
attributes may trump one or more central attributes in our judgments of category
membership. Membership itself, in fact, is not seen to be an all-or-nothing condition.
Instead, categories are graded, meaning that different objects hold different degrees of
membership within a category.
For the study of form in rock music, a prototype-based approach offers numerous
advantages. To begin with, we can realize that the difficulty in coming up with definitions for
section labels resides within the inadequacy of the way a definition-based approach models
human perception. We must assume, for instance, that no definition of a chorus section
exists. Instead, we ascribe chorus-like quality to a span of music based on a confluence of
factors, none of which are in and of themselves necessary to the quality of the section. A
prototype approach helps us appreciate that individual sections evoke form labels in various
degrees of strength or weakness, harmony or dissonance, purity or mixture. Although
theorists will, in all likelihood, continue to say that a particular section “is” or “is not” a verse,
chorus, bridge, etc., a section may instead be viewed as manifesting qualities of one or more
section types more or less prototypically. We can also imagine that by changing various
attributes of a section type, we can transform one category into another (i.e., verse to chorus).
Since section categories display graded membership, moreover, the potential for ambiguity
and blends between section types is relatively high. Ultimately, the recognition that various
attributes contribute to our categorization process in rock music opens up a complex research
area, in which the effects of a host of factors should be considered with regard to their role in
our categorization process.
In its current state, however, form theory in rock offers only limited information as to
what musical effects activate our sense of the role (or roles) a section plays within a song.
Existing descriptions of song sections are universally perfunctory, and few guidelines are
provided as to how to apply these section labels. When describing the verse-chorus form, for
Chapter 1: Prologue
5
example, John Covach mentions that “the focus of the song is squarely on the chorus” (2005,
71), but how this focus is identified, generated, or measured remains unexplained in Covach's
discussion. Theorists also provide many examples of songs to elucidate their section
definitions, but the reader is usually left to infer the details of how and why a particular
passage expresses a given label. When theorists do provide descriptions of section labels, the
emphasis is mostly on lyric structure (Stephenson 2002, 135; Moore 2001, 52; Everett 2009,
145). But formal structures in lyrics and music are often not aligned, and the extent to which
musical factors are involved in our perceptual process is an underdeveloped topic.
The development of a broader and deeper understanding of the prototypical elements
of form in rock music is a central task of this dissertation. I have argued that we must put
front and center in our theoretical system the prototype effects we find latent in theorists’
descriptions of song sections. This dissertation thus aims to lay a more solid foundation for a
prototype-based approach to form in rock music. The hope is that this will allow us to model
more effectively how we hear this music and, perhaps, begin to unravel why it continues to so
greatly sustain our fascination.
Dissertation overview
The following pages are organized into four central chapters, which are framed by
this introductory chapter and an epilogue. In Chapter 2, a context will be created for the
original work in the chapters that follow. This context will be divided into five overviews: 1)
research within the field of cognitive science on theories of categorization – specifically,
research that supports a prototype-based approach, 2) the use of the term “prototype” within
the writings of music theorists, 3) current methodologies of form within the music theory
community as well as the intersection of these methodologies with a prototype-based
approach, 4) work in the field of music cognition on the perception and effects of form in
general, and 5) mainstream approaches to form in rock music. The central task of Chapter 3
will be to develop prototypes for various section labels. These section labels – which will be
referred to as section “roles” (for reasons to be discussed) – will be limited to verse, chorus,
refrain, bridge, solo, prechorus, intro, outro, and link. An attempt to make explicit what is a
complex perceptual process can never hope to achieve complete success. But a much more
detailed understanding of section roles will emerge from this discussion. To come to this
understanding, three basic methods will be employed: 1) the synthesis of criteria offered by
other authors, 2) the identification of other criteria to help explain the more general
observations of other authors, and 3) the presentation of original criteria derived from
personal intuitions. In essence, the main goal of this chapter will be to present and describe
clear cases of section roles. In Chapter 4, we will explore how these section roles interact (or
Chapter 1: Prologue
6
intersect) with different organizational schemes. Organizational schemes are particular types
of melodic and harmonic patterns commonly found in rock music. Three main organizational
schemes will be discussed: 1) the blues (12- and 16-bar), the SRDC pattern (16 bars), and the
AABA pattern (32 bars). As various attributes in these organizational schemes change from
song to song, we can see the schemes “convert” from one section role to another. A central
benefit of examining this conversion process is that we expose potential paths in the historical
development of form within rock music. The final main chapter (Chapter 5) will be devoted
to exposing a particularly interesting subset of ambiguities between section roles. In many
cases, we are confronted with strong evidence for one section role, yet the passage seems to
also be playing the part of another role within the larger context of the song. In this regard, it
appears as if we have blends or amalgamations of section roles. Specific types of blends often
recur in rock songs. On its own, each blend might seem to be an isolated case of ambiguity
between section roles. But the recurrence of particular types of blends in numerous songs
argues for the importance of calling greater attention to such cases.
Some preliminary issues
Before concluding this introductory chapter, it will be helpful to address some
preliminary topics. Specifically, it is worth discussing both the technical aspects of the
musical examples as well as the repertoire under consideration. There are other preliminary
issues that will need to be addressed (such as measure lengths), but these will be reserved for
the introductory portions of later chapters so as not to overwhelm the reader with such
concerns here.
This dissertation uses two primary means to convey information about a song to the
reader: section succession charts and melodic-harmonic transcriptions. A section succession
chart (e.g., Example 3.3.02 in Chapter 3) is used to provide a large-scale overview of the form
of a song. These charts include the approximate start time for various segments of the song
and, typically, the measure lengths of these segments. (When measure lengths are not
included, it is often because these lengths are somewhat ambiguous.) These charts also
usually contain labels for these song segments, information about potential larger-scale
groupings of these segments, as well as useful yet limited glimpses into the attributes of
selected domains (such as lyrics or texture). Section succession charts are often employed in
the writings of music theorists on rock music; they are especially prevalent in the work of
Covach (notably his 2005 and 2009 publications). Although these section succession charts
are helpful tools to convey the “big picture” for a song, they have some inherent
shortcomings.
Chapter 1: Prologue
7
The shortcomings of form charts derive from their basic structure, which is
essentially a table with data cells organized into rows and columns. The nature of distinct
rows implies that there is clear separation between two segments in a song. Of course,
consecutive segments often overlap with one another in musical practice (e.g., the last
downbeat of one section can also be the first downbeat of the following section). These
overlaps are important moments in the form of a song (as we will see) that unfortunately get
lost within a table format. Another shortcoming of these charts is that they inherently
encourage analysts to choose a single label for a particular song segment. In particular, the
small cells do not allow for much more than just a brief label. Despite these shortcomings, no
superior method currently exists for representing (in a visual format) large-scale form in rock
songs. Working within the system, therefore, we should exercise a certain amount of
sympathy to the creator of the form chart. At times, one might feel that an argument put
forth in this dissertation is more of a response to the limitations of the form chart medium
than it is a meaningful disagreement with the analysis of an author. Certainly, the form chart
medium may be seen as the primary culprit in many instances. Nevertheless, we should
exercise caution when constructing these charts, for they are a central resource in how we
understand (and convey our understanding of) form in rock songs.
The bulk of the other examples used in this dissertation are transcriptions. These
transcriptions (e.g., Example 3.2.01 in Chapter 3) capture the basic melodic, harmonic, and
lyric content of a short passage from a song. As a result, these transcriptions are inherently
limited to only a few domains – in particular, those domains that are pitch-based. Many
authors have complained about the emphasis on notation and pitch relationships within the
music theory community, especially with regard to the analysis of rock music (Tagg 1982, 42;
Middleton 1990, 105; McClary and Walser 1990, 281). Undoubtedly, the transcriptions in
this dissertation inherently neglect many parameters – such as timbre, instrumentation, and
dynamics – that fall outside of what transcription itself can easily capture. Even within the
domains of pitch and rhythm, these transcriptions minimize many interesting aspects, such
as rhythmic nuances and microtonal inflections. The purpose of these transcriptions,
however, is not to unduly draw the attention of the reader to only a limited set of attributes.
Instead, the transcriptions are provided primarily to help the reader refer to particular
locations within the form of the song. There can be no substitute for listening directly to the
original recordings, for many of the analytical observations presented in the following pages
are predicated on qualities that cannot be adequately captured via notation. (Web sites such
as youtube.com and grooveshark.com provide easy access to digital versions of the song
examples.)
Chapter 1: Prologue
8
That being said, the transcriptions do act as useful visualizations for certain pitch-
based aspects of a song. We should be aware, however, that they represent analyses of song
segments. (Winkler 1997 describes in detail how the acts of transcription and analysis are
fundamentally inseparable.) This subjectivity – inherent in the notation of recorded sound –
is embraced within this dissertation. For example, the transcriptions herein often omit
background vocals so that the organization of the primary vocal melody can be more clearly
seen. The harmonic textures supporting this main vocal melody have also been reduced to
Roman numerals. As is well known, Roman numeral analyses by different music theorists do
not always agree (see de Clercq and Temperley 2011). The use of Roman numerals thus
represents an interpretation of the harmonic context within a tonal framework. Also, for ease
of reading, each song has been transposed to C (major or minor). (The melodic transcriptions
may thus be considered to represent a scale-degree-based hearing that includes information
on relative registral placement.) Other interpretative elements of these transcriptions include
measure lengths and melodic phrase markings (a more in-depth discussion of these elements
will be reserved for the introductory portions of Chapters 3 and 4, respectively). The purpose
of these transcriptions is clearly one of a scholarly (i.e., non-commercial) nature, and their
use is aimed at the enrichment of the academic community as well as the wider public at
large. Because the transcriptions are analyses and interpretations of only portions of larger
musical works, moreover, they do not supersede either the original recorded versions or
commercial sheet music versions. As a result, the transcriptions in this dissertation are
considered to fall under the purview of “fair use.”
A final preliminary issue to be addressed is the repertoire under consideration. In
particular, the question remains unanswered as to what exactly constitutes “rock music” and
“successful songs.” To begin with, the issue of what constitutes “rock music” has been a topic
with which theorists have wrestled for many years. As discussed in de Clercq and Temperley
2011 (50-1), we can find both broad and narrow conceptions of this category in academic and
journalistic writing. In a narrow sense (e.g., Temperley 2011), rock is a genre of popular
music that is distinct from other closely-related categories, such as pop, soul, or rap. Even
within these narrower conceptions, rock is seen as still encompassing a number of large sub-
genres, such as hard rock, alternative rock, punk rock, and classic rock. We can also find
conceptions of rock that cast a much wider net. In Covach’s textbook on the history of rock
(2009), for instance, he presents examples ranging from Johnny Cash to Madonna to Public
Enemy. Of course, Covach does not explicitly say that all of these artists clearly represent
rock music. In fact, we can safely assume that early examples in his textbook (e.g., those by
Judy Garland and Howlin’ Wolf) are included because they are part of the history of rock, not
necessarily because these songs should be considered central examples of rock music.
Chapter 1: Prologue
9
Undeniably, rock music incorporates stylistic influences from a variety of pre-rock genres,
such as blues, jazz, country, folk, and Tin Pan Alley music. As a result, we could say that the
category of “rock music” evinces prototype effects, in that some examples seem to be more
central than others (the narrower conception), while other examples seem to be less central
(the broader conception). Evidence of asymmetrical category structure is to be expected, in
fact, based on the naturally-developed meaning for the term “rock” (as will be explained in
Chapter 2).
The conception of rock music used in this dissertation is of the broader variety, as
reflected in its musical examples. This broader conception is sympathetic to the approach
used in recent discussions of harmony in rock music (e.g., Everett 2004 and Doll 2007). The
general feeling is that – whether via cross-pollination or some other vehicle – songs across a
large swath of popular music share similar compositional practices. With regard to the study
of form, the similarity in these compositional practices is reflected by that fact that music
theorists use the same section labels (with the same implied meanings) over a wide range of
popular music. Moreover, we find evidence of commercial musicians effortlessly shifting
between what are often considered to be disparate styles. (Consider, for example, the three
versions – pop/rock, country, and world – of the 2002 album “Up!” that were released by
Shania Twain [as discussed in Neal 2008].) It is thus with a relatively inclusive approach to
rock music that any comprehensive discussion of section labels should probably take place.
A prototype approach to the category of “rock music” ties into the “successful” aspect
of the dissertation title as well. “Success” is understood here not as something that is
necessarily determined in commercial or chart performance terms. For instance, the song
“September Gurls” (Big Star, 1974) was never a big seller, yet it ranks within the top 200
greatest songs of all time (according to Rolling Stone magazine [2004]). Success, in other
words, can be measured in a number of ways. While album sales or chart performance are
certainly measures of success, we should also consider other factors – such as critical acclaim,
concert ticket sales, or influence on other musicians – as viable indices of general success.
Years of perspective make the issue of success somewhat more clear, as history tends to be the
greatest arbiter. Nonetheless, certain songs seem to have garnered general success, whether
via commercial sales, critical acclaim, or some other means. This dissertation attempts to
primarily employ such songs in its examples. In doing so, there is no attempt to establish a
canon of rock music (pace Moore 1992 or Hubbs 2008). Rather, there is an attempt to
interact with the unstated canon that already exists within the community of rock musicians,
scholars, listeners, and critics.
10
Chapter 2: Background
2.1: Introduction
This chapter lays out the theoretical context on which the dissertation is based. It
begins with an overview of work on categorization within the field of cognitive psychology,
since the way we categorize things in general presumably parallels how we categorize song
sections in particular. As we will see, this work shows that the way we understand and use
categories cannot be codified in any simple way. One theory that attempts to account for the
complexities of our categorization process involves what have been referred to as
“prototypes.” The term “prototype” requires its own dedicated discussion, though, as it has
had a number of different meanings over the years, both in the field of cognitive psychology
as well as music theory. Consequently, some time will be spent comparing and contrasting
these meanings to better highlight the specific interpretation that is adopted herein.
Following this discussion is a brief overview of some modern music-theoretic approaches to
form. We will find, interestingly, that aspects of many of these approaches can be seen as
responses to the nature of our categorization process. One might assume that the realm of
music cognition would be able to offer some insights into these issues, but unfortunately,
research on the perception of musical form is sparse and of questionable applicability. The
final portion of this chapter summarizes contemporary theories of form in rock music. This
summary is not framed directly in terms of prototypes, per se. Instead, it primarily presents
some terminology and theoretical background that will be useful in the chapters that follow.
It is in these following chapters that the reconciliation of prototypes and form in rock music
will take place.
2.2: Cognitive Science and Categories
The standard section labels in pop/rock music (verse, chorus, etc.) are – on a
fundamental level – categories into which we assign parts of songs. The words and labels we
use represent concepts, and these concepts are essentially mental representations of a
particular class of things or processes in the world (i.e., conceptual categories). In this regard,
cognitive psychology has much to tell us about the mental processes we use in making section
label assignments, since the past few decades have seen much research into aspects of
conceptual categorization. Although our ability to parse songs into various sections may
initially seem like an elementary task, experimental work in cognitive psychology suggests
Chapter 2: Background 11
that the way we learn, apply, and use categories is a rather complex topic. From the 1970s
work of Eleanor Rosch and beyond, researchers have noted asymmetries – i.e., prototype
effects – in the way we judge category membership, in that some members of a category
appear to better represent that category than others. These effects challenge any simple
explanation of the categorization process (musical or otherwise). The following pages provide
an overview of this work. Very little of this research has much to say directly about music. Be
that as it may, it is doubtful that the general principles we see in a variety of other domains
(vision, linguistics, etc.) are not related in a significant way to the specific principles we
observe in music.
The classical approach
Because we often identify things – such as a chair – in a seemingly quick and
effortless manner, it is easy to assume that the process by which we go about this
identification is relatively uncomplicated. Until about 1970, in fact, “the psychology of
concepts” (as this field has been called) was considered to be a fairly straightforward topic.
As a result, cognitive scientists worked for many years under the assumption (whether
implicit or explicit) that we understand the words and labels we use through a definition-
based approach (Murphy 2002, 12). The pervasive use of definitions has such a great history
in Western thought, in fact (tracing all the way back to the formal logic of Aristotle), that
Smith and Medin have dubbed it the classical view of concepts (1981). Others, such as
Lakoff, refer to the use of definitions as objectivism (1987, xii), since it derives from a world
view in which a particular attribute or feature is an integral part of the thing itself.
A definition-based approach indirectly asserts four different yet related aspects of
categories. These aspects can be described as necessity, sufficiency, clarity, and uniformity
(Murphy 2002, 15). The first aspect – necessity – is that a potential category member must
necessarily have the specified attributes of the category in order to belong; otherwise it
cannot be a member. The second aspect – sufficiency – describes the fact that if a potential
member has these specified attributes, then we have sufficient information to accord it
category membership. Clarity relates to the fact that – given the aspects of necessity and
sufficiency – something is either in or not in a category. There can be no ambiguous cases
since a definition does not allow for statements other than those that are true or false.
Finally, the classical view does not make any distinction between category members. If
something meets the requirements of a definition, then it is as good a category member as any
other category member. This final aspect displays the quality of uniformity, in that all
members of the category are equally good examples of the category, and correspondingly,
anything outside the category is an equally bad example.
Chapter 2: Background 12
Problems for the classical view
For better or for worse, the way in which we interact with our environment does not
seem to follow the simple, elegant design described in the classical view. Although we saw in
Chapter 1 that some music theorists use a definition-like approach in their analysis of song
sections, we also saw that there were aspects of other theorists’ explanations that did not
seem definition-like at all. A large body of both theoretical and experimental research shows
that the issue of categorization is, in fact, much more complicated than can be accounted for
by the classical view of category structure.
An early critique of the classical view can be found in the writings of the philosopher
Ludwig Wittgenstein. The commonly-cited example is Wittgenstein’s discussion of the
category “game” (1953). Wittgenstein argues that – although we may assume that every sort
of game has something in common – there is, in fact, no single thing that is common to all.
Some games involve competition between players (tennis), while others do not (“Ring Around
the Rosie”); some games involve multiple players (bridge), while others do not (solitaire);
some games are amusing (charades), while others are not (a mind game). As we go through
the list of things that we think are games, features crop up and disappear, creating an
elaborate network of overlapping similarities. Wittgenstein describes this characteristic of
the category “game” as a family resemblance, since members of a family share many features
but do not all share a single common feature. Family resemblances create problems for the
classical view – particularly with regard to the aspect of necessity – since category members
do not necessarily have to have any one single feature for category membership.
Another issue for the classical view is how to account for the “fuzzy” nature of many
categories (see Zadeh 1965). Categories such as “rich people” or “tall women,” for example,
do not have the aspect of clarity required by a classical approach because there are no clear or
absolute conditions upon which we judge wealth or height. This aspect reveals the graded
nature of our conceptual categories and raises the issue of category boundaries (Lakoff 1987,
21-2). Category membership in these cases is not an either/or situation. Instead, a category
such as “rich person” is one in which some members more clearly belong, while the status of
others is more ambiguous. This last example shows that the categorization process often
involves taking a continuous parameter (e.g., wealth) and turning it into a discrete property
(e.g., “is” or “is not” rich). The external world presents us with a variety of continuously-
variable domains, and language (as a manifestation of the categorization process) divides up
these continuous domains into discrete categories.
The domain of color, for example, is a physical continuum (the frequency of light
waves) that language discretely partitions, and some of the earliest experimental evidence of
prototype effects derives from work on color perception. Research by Berlin and Kay revealed
Chapter 2: Background 13
regularities in the way color was linguistically encoded by different cultures (1969, 1-2). In
particular, these regularities appeared when subjects were asked to identify the best example
of a specific color term. While one culture might have a basic color term covering the range of
both blue and green, the best example for this term would not be turquoise (i.e., a mixture of
blue and green). Instead, a standard focal blue or focal green would be the color chosen to
best represent that color category. Such findings contradict notions of uniformity within a
category, since certain members consistently appear to be better examples than others.
Research reported in Rosch 1973 confirms these results. Consequently, Rosch argued that
our perception of color centers on perceptually-salient “natural prototypes.”
Of course, a large component of color perception relates to physiological factors. But
further research by Rosch and others extended these findings to other domains. In Rosch
1975b, for example, the domain of numbers was shown to exhibit asymmetrical structure.
One study showed that when subjects were given the sentence, “X is virtually Y,” they reliably
placed particular category members (i.e., those that were more prototypical) into the “Y”
position. Statements such as “11 is virtually 10,” for example, were preferred to the opposite
statement, “10 is virtually 11.” (This finding highlights the difference between the
mathematical definition of numbers and our mental representation of numbers. More on
that distinction below.) Similar results were reported in Rips 1975, which found a
relationship between prototypicality and induction. Subjects were more willing, for example,
to assume that birds as a whole would catch a disease if that disease were found in robins
than if the same disease were found in ducks. (Robins are thus more prototypical birds than
are ducks.) Prototypicality also appears to affect the speed with which people make
judgments of category membership. Rips, Shoben, and Smith (1973), for example, found that
subjects were able to categorize a robin as a bird much more quickly than they could
categorize a chicken as a bird. These effects held for visual stimuli as well, in that identifying
a picture of a robin was shown to be faster than identifying a picture of a chicken (Murphy
and Brownell 1985, reported in Murphy 2002, 23). Other asymmetrical aspects of categories
were found to relate to the lack of complete transitivity and hierarchy in conceptual
categories. In studies by Hampton, for example, subjects reported that tree houses are in the
category “dwellings that are not buildings” but also that tree houses are in the category
“buildings” (as reported in Murphy 2002, 26-7). Similarly, subjects judged chairs to be
furniture, car seats to be chairs, but denied that car seats were furniture. A definition-based
approach to categories does not predict the lack of transitivity and hierarchy in these
responses.
Chapter 2: Background 14
Types of categories
Overall, we can see that the qualities of necessity, sufficiency, clarity, and uniformity
– all of which underlie a classical approach to categories – do not hold up well under scrutiny.
Before continuing the discussion, however, it is worth clarifying what is meant by the term
“category” itself. Not all categories are equal, and the skeptical reader has probably tried to
imagine various counterexamples to those provided above.
Rosch uses the term “natural prototypes” because she is attempting to describe the
cognitive mechanisms behind “natural categories.” We thus need to care to distinguish
between natural categories and artificial categories. Artificial categories are those that have
been consciously planned, devised, or determined by humans themselves. In essence,
artificial categories are definitions. Such categories appear, for example, in rules,
mathematical logic, and experimental designs. The category of “even numbers,” for example,
is mathematically defined to be the category that includes integers divisible by two without a
remainder. Yet our understanding of even numbers may not be so straightforward. For
example, the number 444444443 seems to be “more even” than the number 375193713, even
though both are mathematically defined to be equivalently even (see the discussion in Lakoff
1987, 148-51). We therefore understand numbers as natural categories as well. Natural
categories arise in language through no conscious decision of an individual but rather as a
result of our interaction with our environment. As a result, natural categories often have an
unlimited amount of information that can be learned about them. The natural category of
“dog,” for example, potentially contains information about dog behavior, dog diseases, dog
anatomy, etc., not just what is or is not a “dog.”
A central question here is to which category type do section labels such as verse,
chorus, and bridge belong. Or, more broadly, what type of categories should music theory
concern itself with? To a certain extent, music theory is indeed a constructed language that
has been consciously determined by its practitioners. We may define a Neapolitan sixth
chord, for example, as a particular collection of pitch classes with respect to a particular tonal
center. Then again, certain concepts from music theory are notoriously fuzzy, such as the
distinction between tonicization and modulation. The difficulty that theorists have had in
attempting to define song section categories points to the naturally-developed meaning of
these categories. Qualifying words such as “usually” or “often” show evidence of the
asymmetrical way in which we understand song section labels, since not all category members
appear to be equal. The fact that section labels stand as natural categories can be seen to
derive from the origins of these terms within the vernacular tradition of the popular music
community (i.e., songwriters, performers, critics, and listeners). A classical, definition-based
Chapter 2: Background 15
approach cannot be considered to reflect our shared understanding, therefore, since – as we
have seen – a classical approach fails to model natural categories very well.
Theories of conceptual categorization
While we find strong evidence of asymmetrical membership in our categorization
process, it is not entirely clear as to how we should account for this evidence. The
experimental phenomena themselves do not necessarily imply a particular theory of mental
representation, and so the question remains as to how exactly we process and represent
category information. A few theories have been presented in the field of cognitive psychology
in this regard, three of which will be discussed here: the prototype model, the exemplar
model, and the knowledge model.
The first view – the prototype model – posits that the observed effects derive from
the existence of prototypes themselves. But what is a prototype? Sometimes, writers state
that a prototype is the “best example” of a category (Lakoff 1987, 7; Murphy 2002, 28).
Statements such as these potentially mislead the reader into thinking that there is a single
best example for a category. In some cases, of course, there is a single best example. For
instance, the focal color red is the central member of the category “red.” Unfortunately, many
readers of Rosch’s work (especially those that were familiar with only her early work)
interpreted her research to imply that there is a single prototype or best example for every
category (Murphy 2002, 41). Yet while a single best member may exist in some categories,
other categories do not appear to be represented this way. For example, what would an ideal
dog look like? More importantly, how could this ideal dog represent the fact that some dogs
are black, large, long-nosed, and short-haired while others are beige, small, snub-nosed, and
curly-haired? One central issue is that a single ideal type gives no information about the
amount of variation that exists within a category (Murphy 2002, 42). Some categories are
relatively wide, while others are relatively narrow (e.g., compare the large variability in types
of pet dogs to the small variability in the types of pet cats). A single best example does not
capture the extent of this variability.
As a result, the “single best example” idea has not been widely adopted. Instead, the
central tenet of prototype-based approaches has been their reliance on features (or
attributes). One such view has been referred to as the summary representation model, which
is illustrated in a statement by Zbikowski: “If the values of the prototype for bird were small,
brown, chirps, and flies, then a wren would be most typical of the category” (2002, 42).
“The” prototype that Zbikowski describes is obviously not the best example. Instead, it is a
list of attributes. This view posits that natural categories are cognitively represented via some
collection of features (Murphy 2002, 42). These features are weighted (some are more
Chapter 2: Background 16
important than others), and this collection describes the category as a whole. Note that
because this collection of features is basically just a list, it can include contradictory features.
A summary representation of a bird, for example, might include a high weighting for “brown
feathers” and a lower weighting for “red feathers.” Prototype-based approaches do not always
have to represent the category as a whole, however. The features may also be correlated in
some way or describe subtypes of the category. For example, some researchers have
presented structured representation models, which consist of dimensions (or slots) and
values in those dimensions (fillers) (Murphy 2002, 47). Each dimension restricts the type of
values it may contain, and thus concurrent contradictory values are avoided. If a bird is
flightless, for example, it would also be associated with not flying south for the winter.
In this dissertation, a prototype-based approach will be the primary mode of handling
categories of form in rock music. A “prototype,” in other words, captures the attributes of
those members that are judged to be more central to that category. In some cases, this
prototype involves a summary of the category as a whole, while in others it involves a
structured representation of a particular class of members. In either situation, our
understanding of the category is seen to strongly derive from the attributes themselves.
One important aspect to keep in mind when discussing attributes is the difference
between frequency and typicality. In particular, the frequency of one type of member within a
category is not necessarily directly linked to its typicality. It is undoubtedly true that
frequency (or “norms”) and typicality are related; part of the prediction process must involve
knowing which features are more likely than others. But atypical members may, in fact, be
very common. For example, chickens are a type of bird that is very frequently discussed,
thought about, eaten, and seen (at least in pictures). Yet compared to a robin (which is much
less frequently discussed, thought about, etc.), a chicken ranks as a much less typical member
of the bird category (Murphy 2002, 31). It appears, therefore, that our conceptual categories
are not necessarily structured in terms of statistical distributions. For a theory of form, the
good news is that typicality effects for natural categories are fairly consistent across
individuals. In other words, research shows that subjects overwhelmingly agree in their
judgments of what are good examples of a category, even when these same subjects disagree
as to the exact boundaries (Rosch 1978). This insight is crucial because it implies that – even
though we as theorists may disagree on the categorization of certain borderline musical
examples – we should be able to find common ground with regard to the category in general.
Although a prototype-based approach explains a great deal of the experimental
evidence on conceptual categorization, other theories have also been proposed in its stead.
One of these is the exemplar model (derived from Medin and Schaffer 1978). The basic idea
with this view is that subjects make classification judgments by retrieving stored exemplars,
Chapter 2: Background 17
i.e., specific remembered instances that have been previously encountered. Another model of
conceptual categorization is the knowledge approach (also known as the “theory theory”).
Basically, this view holds that we use knowledge, reasoning, and theories to decide whether a
particular example should be in an existing category or some new category (Murphy 2002,
61). Overall, no single theory currently exists that can account for all of the results found in
experimental studies on categorization. But this does not mean that these theories are
necessarily incompatible. Rather, people may use different strategies for different tasks.
Subjects have been shown, in fact, to switch between a prototype approach and an exemplar
strategy within the same experiment (Smith and Minda 1998). We may even be able to
integrate various approaches within the same categorization task. Presenting these theories
as one versus another may thus be somewhat misleading. Nevertheless, a prototype-based
approach has provided some of the best answers to date on the psychology of concepts, and
consequently, this approach will serve as the backbone of the work herein.
2.3: Prototypes and Music Theory
The term “prototype” has been employed within music theory at various points
throughout the history of the field. But this theoretical work has not always involved similar
methodologies. Of course, the term “prototype” existed long before cognitive psychologists
endowed any particular meaning upon it. The prefix “proto-” can mean “first in time,”
“original,” or “primitive.” Thus a music theorist might simply use the word “prototype” in the
vernacular sense – as referring to an underlying pattern. This situation appears to be the case
with the concept of metric prototypes, for example, as found in Schenkerian theory (see
William Rothstein [1989] and Carl Schachter [1987]). Schenker’s original term for this
concept – metrisches Vorbild (1935) – might just as easily have been translated as “metric
model” or “metric example.” Theories of music do not necessarily have the same goals as do
theories of mental representation, so we should not expect usages to be the same between the
two disciplines. Often, though, music theorists do use the term “prototype” with an explicit
reference to work in cognitive psychology. A few of these usages are discussed below, which
should help distinguish between various shades of meaning this term has held.
Approaches to prototypes
Sometimes, theorists adopt the “single best example” approach to prototypes. For
instance, Agmon 1995 re-conceptualizes functional harmonic theory in terms of prototypes
and categories. In this article, Agmon specifically mentions psychological research in the
color domain (Harnad 1987), where each basic color category has a single focal prototype.
Chapter 2: Background 18
The functions of tonic, subdominant, and dominant are seen as cognitive categories, for
which the pitch collections represented by I, IV, and V respectively stand as the prototypes.
Other chords are viewed in terms of “degrees of triadic similarity,” which Agmon defines as
the number of common tones a triad has with these prototypes (199). With respect to I, for
example, certain chords are seen as maximally similar (III and VI), others as intermediately
similar (IV and V), and yet others as minimally similar (II and VII). This conceptualization
captures the graded membership of natural categories as well as the fuzzy nature of functional
labels, in that certain triads may have dual citizenship (e.g., VI is maximally similar to both
tonic and subdominant functions).
By adopting a “single ideal type” approach, Agmon limits the explanatory depth of
these category labels. As Agmon himself admits, he separates out a theory of functions from a
theory of chord progression. The standard T–S–D–T paradigm of functional succession is, in
fact, not within the scope of his prototype theory (204). Yet part of our notion of dominant
function certainly relates to its strong tendency (at least within the context of common-
practice-era music) to move to tonic. Agmon, therefore, does not construct his theory of
functions as an attempt to broadly model our understanding of these function categories.
Rather, Agmon has specific goals for music theory – in particular, creating a context for non-
standard usages of function labels. For instance, an apparent II chord may (albeit weakly) act
in a dominant function, such as is found in the opening bar of Haydn’s Piano Sonata in D
major, Hob. XVI: 37 (which alternates I and II chords).
Another setting within which we find the term “prototype” employed is within the
writings of Matthew Brown (e.g., 2004 and 2005). In his work, we find Schenkerian theory
re-formulated through the lens of cognitive psychology. As is well known, Schenker posits
that every tonal composition can be derived from a single background model – the Ursatz –
of which there are three possible configurations (1935). Each Ursatz is a soprano-bass
framework that embellishes a simple tonic-dominant-tonic progression, the only difference
being the starting soprano scale degree (^8, ^5, or ^3). Instead of referring to these three
Ursätze by their common translation (“fundamental structures”), though, Brown calls them
“prototypes” (2005, 66). These prototypes, according to Brown, capture the essence of tonal
voice leading and harmony. The great composers internalized these prototypes and
generated masterworks via the application of a recursive set of transformations (2004, 155).
It is not immediately obvious from Brown’s discussion what natural category (or
categories) these prototypes might reflect. In fact, it does not appear that Brown is
necessarily using prototypes to explain prototype effects – such as graded membership or
unclear boundaries – within any particular category. We might surmise (especially on
reading earlier work such as Brown et al. 1997) that the category accounted for by these
Chapter 2: Background 19
prototypes is tonal music itself. Yet that does not seem to be Brown’s primary purpose here.
Instead, Brown’s prototypes represent internalized musical patterns. His prototypes are
highly-structured mental models (whether conscious or unconscious) that capture
information within a limited domain (pitch). Brown is thus positing a specific theory of
mental representation, one that many have found useful to conceptualize the underlying
melodic-harmonic structure of common-practice-era compositions.
One other work in which prototypes play a central role is Robert Gjerdingen’s 2007
book, Music in the Galant Style. In this book, Gjerdingen presents a conceptualization of
music that is strongly couched in categorization theory. Gjerdingen’s primary purpose is to
expose and explain various schemata that, he argues, are important structural conventions
for mid-1700s music. What is a schema, though? According to Gjerdingen, a schema is the
unspecified mental process or structure that we associate with a conceptual category (11).
Gjerdingen goes on to present three contemporary approaches to the understanding of a
schema – prototypes, exemplars, and theories – which map to those approaches discussed in
the previous portion of this chapter.
In his explication of schemata, Gjerdingen relies on prototypes (and he specifically
calls them prototypes). For example, the prototype for the Ponte schema involves a pedal on
^5 in the bass, above which the soprano voice outlines a dominant chord via its arpeggiation
through ^5, ^7, and ^2 (461). But this prototype is not merely a soprano-bass framework.
Gjerdingen also describes the metrical context in which this scale-degree structure typically
occurs, along with central features and common variants of the schema. Additionally, he
includes a discussion of the typical locations for this scale-degree structure within the form of
a piece, to what common key area the structure leads, and general strategies for its use within
a specific historical timeframe. All of these factors – the scale-degree network, the feature
list, the typical usages, etc. – are wrapped up into the prototype for this schema. For
Gjerdingen, therefore, the prototype is a representation that contains both structured (e.g.,
the scale-degree network) as well as unstructured (e.g., the feature list) components.
Moreover, many of these components are conveyed through abstractions – such as the use of
scale degrees instead of specific pitches – in order to most clearly convey the category to other
music theorists. Thus while the objects of inquiry in Gjerdingen’s work differ from those in
this dissertation, the basic methodology can be seen to be the same. In essence, his
prototypes capture the attributes (whether structured or unstructured) of those musical
examples that are judged to best represent that category (or schema).
Chapter 2: Background 20
2.4: Music Theory and Form
For the most part, the term “prototype” is not employed in the writing of music
theorists. Yet even though this term may be absent, we sometimes find that the approach
taken by a theorist is in sympathy with a prototype-based approach. Contemporary
conceptions of form in common-practice music, for example, can be viewed as different
responses to prototypes effects found in form categories. Even though common-practice
music may be far removed from rock music (stylistically speaking), a brief investigation into
issues of form within common-practice music may benefit a study of form in rock.
One general issue is what M. Evan Bonds has called the “paradox” of musical form
(1991, 13). In short, the term “form” encompasses two contrasting senses. In one sense, form
is the explication of what certain musical works have in common. Theorists search for
patterns among various works and then create abstract models to capture these similarities.
In the opposite sense, form describes the unique structure of a single piece of music.
Theorists want to be able to say why a particular work is in a particular structure. The term
“form” is thus somewhat polysemic in nature, in that it appears to have multiple ways of
being understood. These opposite connotations inherently create tension within any single
perspective to form. In fact, we may categorize various perspectives to form based on the
extent to which they lie on one side of this paradox or the other.
One way to characterize this paradox is as the pull between a “top-down” and a
“bottom-up” approach. (This duality is somewhat similar to the “conformational” and
“generative” approaches described in Bonds 1991 but framed in a more general way.) A top-
down approach to form searches for similarities among a large number of works. These
similarities are used to theorize abstract, ideal types, which then become reference points in
analysis. Specifically, the analyst measures how much a particular piece exemplifies a given
abstract type. Value judgments are typically made on this basis. The artistic merit of a
particular piece, for instance, is often seen to directly relate to those unique features that set it
apart from the more typical manifestations of the model. What is exceptional is seen to
valuable. In contrast, a bottom-up approach views the form of a piece primarily as the
consequence of the content itself. The final form of the work is seen to result from the
combination of various elements. In this regard, the bottom-up view inextricably intertwines
form and content. Via this method, the artistic merit of a given piece relates to how closely
lower-level units (e.g., motives) can be seen to determine higher-level units (phrases and
sections). Unity, cohesion, and organicism are valued in this view.
The paradox to form could be framed in yet an even more basic way. Specifically, this
paradox can be seen to derive from the fundamental divide between theory and analysis. To
Chapter 2: Background 21
paraphrase Arnold Whittall: theory generalizes, whereas analysis particularizes (2010). In
other words, theory explains what kind of similarities we can expect to find in a body of
music, while analysis unpacks this information within the context of a specific work. The two
opposing meanings of form can thus be seen to reflect the two different roles that form plays
within theory and analysis.
Modern methodologies
The changing tides of preference between top-down and bottom-up approaches (or
theory and analysis) can be seen in contemporary approaches to form as well. William
Caplin, James Hepokoski, and James Webster each discuss form in their recent book,
Musical Form, Forms, Formenlehre (2009; edited by Pieter Bergé), and their discussion
nicely encapsulates modern methodologies. More importantly, the work of these three music
theorists can be viewed as varying responses to the prototype-based nature of form
categories.
Webster presents an analytical method he calls “multivalence” (in Bergé 2009, 128).
The multivalent approach takes for granted that a musical work encompasses multiple
domains: tonality, rhythm, dynamics, instrumentation, register, etc. Each of these domains
generates its own temporal patterns. At times, the patterns in these domains may be
congruent; at other times, the patterns may conflict. When a particular form can be said to
exist, writes Webster, it is the result of many factors that intangibly contribute to our final
choice of a form label (129). Often, no clear form can be said to exist. In this view, the value
of a great work is seen to directly relate to its “multifariousness.”
This description of form resonates strongly with a prototype approach. Notably,
Webster describes the attribute-rich quality of real-world objects. For Webster, music and
form are multi-dimensional, in that passages of a musical composition often interact with
multiple form types at once. The implication, therefore, is that our conception of a particular
form type (or category) encompasses multiple domains, and our perception of these form
categories may be triggered by only a partial set of attributes. Webster’s multivalent analytic
approach thus highlights the disconnect between the apparent simplicity of a category label
and the complexity of our concept for that category.
In contrast, Hepokoski offers an approach grounded more firmly in theory-building.
Based on his work with Warren Darcy (2006), Hepokoski describes his “Sonata Theory” as
“dialogic” form. This approach posits that an individual composition should be
contextualized within the culturally-available normative gestures that existed at the time the
work was composed. Instead of providing rules concerning the parts of a sonata, Hepokoski
and Darcy expose a set of common options or defaults. When these normative gestures are
Chapter 2: Background 22
denied or eschewed (a “deformation”), an interpretation is demanded. It is the composer that
puts his or her work in dialogue with the conventions, and we are meant to listen in on this
conversation. This approach can be seen to utilize top-down principles. Reference points are
derived from the repertoire, and individual pieces are compared to these reference points.
The approach in Hepokoski and Darcy departs from a typical top-down approach, however, in
the extensive depth to which these authors specify the points of reference. Most notably,
there is no single overarching sonata form. Instead, there are five different types, all of which
have further subtypes. As well, there is no strict line between what is normative and what is
not normative. Rather, there are hierarchies of default strategies, and these defaults adhere
to specific contexts. The analyst who becomes an expert in Hepokoski and Darcy’s system has
not merely memorized a few abstract models. Rather, to understand Sonata Theory is to have
a complex awareness of the various possibilities that can occur within a sonata form
movement.
This elaborate account of sonata form aligns closely with research into categorization.
Overall, Hepokoski and Darcy argue that a single form category cannot be understood
through a single best example or ideal type. There is a conscientious attempt to avoid
defining sonata form too closely, and the question of whether a particular piece “is” or “is not”
in sonata form is one that they specifically rebuff (in Bergé 2009, 75). In creating their
theoretical system, therefore, Hepokoski and Darcy model the complex mental representation
of the sonata form category in the mind of an educated listener.
William Caplin presents yet another approach, that of formal functions. In essence,
formal functions describe the role played by a particular segment of music within the larger
whole. The role of this segment is intimately tied to its temporal location, and so formal
functions designate whether a musical chunk acts as a beginning, middle, or end. For
example, the 8-bar sentence is a formal type in which Caplin posits three formal functions:
presentation, continuation, and cadential. Each of these formal functions has its own salient
characteristics that assist the listeners in locating themselves within the flow of the music.
Continuation function, for instance, is seen to be generated via four compositional devices –
fragmentation, harmonic acceleration, increased rhythmic activity, and sequential harmonies
– though none of these qualities is a necessary condition of the function (1998, 40). Caplin
uses a similar approach to describe formal functions for a variety of other formal types, such
as the period (antecedent and consequent functions) and small ternary (exposition,
contrasting middle, and recapitulation functions). Distinguishing between formal functions
and formal types is a central strength of Caplin’s theory. It allows for an easy description of
hybrid forms, which begin in the manner of one form type but end in the manner of another.
As well, functions may fuse. Thus the second half of an 8-bar sentence typically combines
Chapter 2: Background 23
continuation and cadential functions into a single four-measure phrase. A single formal
function may also be encompassed by numerous iterations of form types, such as when the
cadential function of a subordinate theme is expanded to embrace multiple phrases (in Bergé
2009, 58).
In Caplin’s approach, we see a number of concepts related to categorization theory.
Caplin posits prototypical musical structures – e.g., the sentence – and argues that our
perception of non-clear members derives from these clear examples. Caplin does not pass
aesthetic judgment on whether something is or is not a clear member of a category. Instead,
he uses membership to gauge what a particular span of music might be communicating to a
listener in terms of its role within the work. In this regard, Caplin’s approach overlaps a great
deal with that taken in this dissertation, as we will see in the following chapters.
2.5: Music Cognition and Form
When music theorists enter into passionate debates on a topic – such as that seen
between Webster, Hepokoski, and Caplin – it is often useful to balance these arguments with
cooler observations. The field of music cognition has been helpful in this regard, as it can
provide – through experimental research, computational modeling, or corpus study –
objective answers to some of the questions that a theorist might ask. Unfortunately, existing
work within the field of music cognition on the subject of musical form is rather limited, and
the applicability of this work to form in rock music is unclear. A brief overview of this
research is warranted, nonetheless, if only to show these limitations. In general, these studies
investigate what sort of effects the re-ordering of musical units has on a listener.
Experiments in form manipulation
Most extant experiments involving the manipulation of form have failed to show that
structural changes have any significant effects on listeners’ responses. In Karno and Konečni
1992, for example, the authors tested the effects of structural changes to the form of the first
movement from Mozart’s Symphony in G minor (K. 550). Listeners evaluated – on scales of
“pleasingness,” “interestingness,” “desire to own,” and “best overall structure” – both an
original version of the piece as well as four altered versions, which were created by re-
ordering large chunks of the movement (e.g., first theme, transition, second theme). Yet no
clear preference for the original version was found; the rearranged versions were rated as
highly as the original. Similar results were found in earlier experimental work (e.g., Konečni
1984; Gotlieb and Konečni 1985). Tillman and Bigand (1996) also report a lack of change in
aesthetic responses between original versions of three piano pieces (by Bach, Mozart, and
Chapter 2: Background 24
Schoenberg) and “backwards” versions of these pieces, in which six-second chunks were
linked in reverse order.
It should be noted, however, that most of the subjects in these studies did not have
extensive experience in the musical styles in which their responses were being tested. In
Tillman and Bigand 1996, for example, more than half the subjects had no musical
experience, and the remaining subjects had relatively little instrumental practice (less than 5
years). Similarly, of the 53 subjects used in Karno and Konečni 1992, only one was familiar
with the symphony. It may not be surprising, therefore, that the listeners were not attuned to
aspects of structure above the phrase level since these listeners were probably not highly
attuned to the style as a whole.
In general, existing music cognition research on the perceptibility of form is far too
preliminary to allow any hard conclusions to be drawn. The failure of these experiments to
show any significant aesthetic effects may simply reflect a failure of experimental design more
than anything else. Notably, the relatively low familiarity of the subjects with the style of
music under study may have been a considerable impediment. We should also be highly
circumspect of how these results might relate to form in rock music. Rock music, for
instance, has additional structural cues (e.g., lyrics) that are not present in the instrumental
common-practice pieces chosen for these experiments. Moreover, the average person has a
much higher exposure to (and thus familiarity of) rock music than classical styles. One
inference we might draw is that small-scale and large-scale structures may not necessarily
induce the same aesthetic effects in listeners. In Tillman and Bigand 1996 (discussed above),
for example, the internal characteristics of musical segments appeared to much more strongly
determine aesthetic response than any particular ordering of these segments. We thus might
take care to differentiate between lower-level factors (i.e., internal section quality) and
higher-level factors (i.e., succession patterns) in our studies of form.
2.6: Form, Music Theory, and Rock Music
As the final portion of this chapter, a brief summary of current approaches to form in
rock music is presented below. This summary will help situate the chapters that follow within
the context of modern theoretical methodologies. These methodologies will be discussed
without any extensive commentary, since the goal here is only to create a baseline for the
following pages. A few general issues will emerge, which will be discussed at the end.
Current theories of rock music generally use a two-stage approach in their
explanations of form. First, the section categories themselves – verse, chorus, bridge, etc. –
are described. Second, the standard succession patterns in which these sections occur are
Chapter 2: Background 25
discussed. In this regard, section quality is presented as somewhat independent of the
particular succession pattern in which it participates. This separation of section quality from
a particular succession pattern is not always possible, though, as we will see. Nevertheless, to
best summarize extant theories of form in pop/rock music, it is most efficient to adopt a
similar format as the original authors. The discussion below thus presents some standard
descriptions of song section types followed by some standard descriptions of organizational
patterns for these section types. The writings of Walter Everett, Ken Stephenson, and John
Covach are used as representative accounts in this overview since – having each published a
textbook dealing solely with rock music (2009, 2002, 2009 respectively) – these authors can
be considered central figures in the current state of rock music theory.
A brief overview of song sections
The various qualities that affect our perception of song section categories will be
detailed extensively in the next chapter. As one of its central methodologies, this upcoming
chapter will digest extant descriptions and usages of section role labels as found in the work
of music theorists. Consequently, a complete discussion of sections labels will not be
presented here. In order to discuss descriptions of organizational patterns, however, it is
helpful to have a basic understanding of a few section labels. One should be aware that only a
small number of section labels are used in standard large-scale categorization schemes. In
particular, the large-scale form types discussed below are contingent on only three section
types: verse, chorus, and bridge. As a result, an understanding of only these three section
types is necessary for a discussion of current approaches to form in rock music. Given below
are glossary entries for verse, chorus, and bridge as found in the work of Covach, Everett, and
Stephenson. The use of glossary entries here is not meant to oversimplify or distort each
theorist’s understanding of these sections. Certainly, more nuanced usages can be found in
their writings. These glossary entries simply serve as a starting point for the reader.
Chapter 2: Background 26
Verse
(Covach 2009): “A section that most often features new lyrics with each repetition within a song, unlike a chorus, which tends to repeat the same lyrics with each recurrence. The verse is often used to tell a story or describe a situation.”
(Everett 2001): “A song’s section equivalent to the stanza, usually placed directly after
any introduction, that nearly always appears with two or three (or, rarely, more) different sets of lyrics, but in rare early cases has one set only.”
(Stephenson 2002): “A section of a song that recurs a number of times with a
different text every (or nearly every) time.”
Chorus
(Covach 2009): “Usually the most important or easily remembered section of a song, containing the title and the catchiest musical material. Not all songs have a chorus, but when one is present, it is the focus of the song.”
(Everett 2001): “A song’s section, nearly always affirming tonic, usually appearing in
the song’s interior, with lyrics that remain constant with each hearing.” (Stephenson 2002): “A musical section that recurs numerous times with a fixed text
of several lines.” Bridge (Covach 2009): “A section in a song that provides contrast to other, more salient
sections of the same song, such as the verse or the chorus. While bridge sections can be quite interesting musically, they are almost never the focal section of a song.”
(Everett 2001): “A song’s contrasting middle section, often beginning in an area other
than tonic and usually leading to a dominant retransition.” (Stephenson 2002): “A passage, usually introduced after the second chorus, used to
lead back to the verse or the final repetitions of the chorus.”
It may be helpful to remind the reader that – while the discussion below summarizes
the work of these authors with regard to various categorization systems for entire songs – the
assignment of entire songs into categories is not the primary goal of this dissertation. As one
reason, these large-scale form categories receive much less currency in discussions of rock
form than do section labels themselves. While we will find many overlaps in the large-scale
categorization schemes that these authors present, there are also clear differences in both
terminology and meaning. One might say, in fact, that the categories into which these
authors assign entire songs reflect artificial rather than natural categories, in that each author
essentially develops his own unique system. In contrast, the same basic set of section role
labels is used by all, and so it is these section categories that have more clearly developed a
shared understanding (and are thus amenable to a prototype-based approach). Nevertheless,
Chapter 2: Background 27
it is important to discuss large-scale form categories, for the way an author categorizes the
form of an entire song can shine a light on the way that author understands the section roles
that participate in them.
Form in the writing of Ken Stephenson
Ken Stephenson presents four large-scale form types for rock music in his 2002 book.
These types are the strophic, rounded binary, verse-chorus-bridge, and compound binary
forms. The first category – strophic form – includes songs that contain two or more
repetitions of the same musical material (139). In other words, strophic songs display a
section succession pattern in which there is only a single letter (e.g., AAA). Stephenson
further subdivides this category into one-part and two-part strophic forms. This subdivision
is based entirely on harmonic factors. A one-part strophic form repeats a single harmonic
pattern throughout. Thus songs in which both a verse and chorus appear over the same
harmonic pattern (e.g., “Bad, Bad Leroy Brown” [Jim Croce, 1973]) are categorized as one-
part strophic forms. In a two-part strophic form, there are two harmonically-contrasting
parts, which represent separate verse and chorus sections. The two-part strophic form is still
categorized as strophic because the two sections (verse and chorus) group together to create
one large repeating unit. Within a top-level AAA pattern, for example, each “A” would
represent a verse-chorus or chorus-verse unit.
Stephenson’s second main form type is the rounded binary (140). Like the two-part
strophic form, the rounded binary also contains two section types. But instead of verse and
chorus, the rounded binary contains either verse and bridge or chorus and bridge. The term
“rounded” is used because the bridge (which is implicitly not the opening section) must
always return to the opening section type. From this description, we can surmise that
rounded binary songs display a variety of possible letter sequence patterns, such as ABA,
AABA, AABAA, AAABA, or AABABA, where “A” represents either a verse or chorus and “B”
represents a bridge. Unlike the strophic song, each letter in this pattern can refer to only a
single section type. In other words, “A” cannot represent a verse-chorus unit as it did in the
single-section strophic form. Stephenson thus makes no particular pattern of sections
explicit, nor do his examples of rounded binary form display any consistent succession
pattern.
The third form category Stephenson describes is the verse-chorus-bridge form, which
plainly designates the section roles it involves by name (140). Stephenson presents the verse-
chorus-bridge form as a hybrid of the strophic and the rounded-binary forms. Like the two-
part strophic form, there exist two distinct sections – verse and chorus – that repeat a
number of times. Like the rounded binary form, however, there is a bridge section that
Chapter 2: Background 28
occurs after two or three of these verse-chorus iterations. Stephenson describes two varieties
of the verse-chorus-bridge form, one in which the bridge leads to a return of the chorus only
and a second in which the bridge leads to yet another verse-chorus pair (141). Otherwise,
Stephenson is not particular with regard to the ordering of song sections. In fact, one verse-
chorus-bridge example he provides (Carly Simon’s “Haven’t Got Time for the Pain” [1974])
has a section succession pattern of VCBCVBC.
The final form type Stephenson discusses is compound binary (141). Songs of this
type divide into two seemingly unrelated halves. According to Stephenson, the first half often
follows one of the previous three form types (thus the “compound” label), while the second
half presents new material. This new material is typically structured around a repeating
melodic line or chord progression. Examples of this form type include “Hey Jude” (The
Beatles, 1968) and “Layla” (Derek and the Dominos, 1970). Stephenson admits that songs in
this category often have their own idiosyncratic structure, though, so this category may
simply serve to account for irregular song forms. For instance, Stephenson categorizes
“Stairway to Heaven” (Led Zeppelin, 1971) as compound binary, yet to say that this song
unambiguously divides into two (and only two) distinct sections is not beyond dispute.
Overall, the form types that Stephenson offers do not appear to be dependent on any
particular succession pattern. Instead, his form types are based exclusively on the section
roles they contain. Songs in strophic form, for instance, contain verse and/or chorus material
only (no bridge). In contrast, rounded binary songs contain a bridge as well as either verse or
chorus sections (but not both). Finally, verse-chorus-bridge songs contain all three section
types. Standing somewhat outside this categorization scheme, the compound binary form
captures the more unique structures found in rock music.
Form in the writing of John Covach
John Covach presents his own categorization scheme for large-scale form types in his
2005 essay, “Form In Rock Music: A Primer.” In this essay, he offers five basic form types:
simple verse, simple verse-chorus, contrasting verse-chorus, AABA, and compound form.
These five form types can be seen to overlap with Stephenson’s four form types, although
some differences are worth noting.
Covach’s categories of simple verse, simple verse-chorus, and contrasting verse-
chorus all fall within Stephenson’s category of strophic forms. In Covach’s usage, simple
verse form refers to songs in which only a single block of musical material is repeated
throughout the song. In simple verse-chorus form, the song consists only of verse and chorus
sections built over the same harmonic material. Finally, a song in contrasting verse-chorus
form consists only of verse and chorus sections built over different harmonic material.
Chapter 2: Background 29
Covach’s contrasting verse-chorus form thus maps to Stephenson’s two-part strophic form.
Similarly, songs in Covach’s simple verse-chorus form would be included in Stephenson’s
one-part strophic form.
Other mappings are not so easy. A song that Covach would categorize as simple verse
would be categorized by Stephenson as a one-part strophic form. In Covach’s system, though,
if a song has only a single repeating section, the analyst would view this repeating section only
as a verse. In contrast, Stephenson’s system allows for a single repeating section to be either
a verse or a chorus. We can thus see some effects of how large-scale formal types potentially
affect designations for the section labels themselves. For Covach, a chorus only exists in
relationship to a verse, and thus a chorus cannot exist as a standalone section type.
The fourth form type Covach offers is the AABA form, which is similar to
Stephenson’s rounded binary. In an AABA form, verse material – the “A” part – is contrasted
by bridge material – the “B” part. Unlike Stephenson’s rounded binary, Covach’s AABA form
specifies a particular succession pattern (the AABA sequence itself). In his analyses, in fact,
Covach views songs that depart from this AABA sequence as examples of “formal
innovations” (2006, 44). One should also note that – like the simple verse form – the AABA
form affects section designations. In Covach’s view, an AABA form consists of only verse and
bridge material. This approach differs from Stephenson’s, as his rounded-binary form may
contain either verse or chorus material for the “A” section. Again, Covach does not allow for a
standalone chorus section without a complementary verse.
Also note the difference in Covach’s system between the mechanisms that determine
AABA form as opposed to those that determine verse-chorus form. For AABA, the form type
is determined primarily through the succession pattern of the parts, and section labels derive
from this pattern. For verse-chorus forms, in contrast, the section labels themselves
primarily determine the form type and no particular succession pattern is implied. This
system potentially creates a tricky situation for the analyst faced with choosing between these
two form types, as it is not clear whether we should emphasize section quality or section
succession until after the form type itself has been identified.
In Covach’s final form type – the compound form – we find an emphasis on both
section type and succession pattern. By virtue of the word “compound,” this form type
appears to map to Stephenson’s compound-binary form. Yet there is not complete equality
between the two. For example, Covach analyzes “Hey Jude” (which Stephenson views as
compound binary) as an AABA form with a coda (2005, 71). Like Stephenson, though,
Covach does use his compound form category to include songs for which no other form type
seems appropriate. Thus Covach sees “A Day in the Life” (The Beatles, 1967) as a compound
Chapter 2: Background 30
ABA form (2006, 48). Yet this song lacks the large-scale AB pattern at the core of
Stephenson’s compound binary form type.
In practice, the term “compound” in Covach’s system is used most often to describe a
particular form type: the compound AABA form. The compound AABA form consists of verse
and chorus material within each “A” part and a bridge section for the “B” part; this form type
thus maps to Stephenson’s verse-chorus-bridge form. Once again, however, Covach’s system
more strongly specifies a particular succession pattern. While Stephenson admits that certain
configurations of verse-chorus-bridge songs are more common than others, Covach views
departures from the compound AABA structure as valuable and “innovative” moments (2003,
185).
Covach thus presents a theoretical system in which AABA patterns (whether
compound or non-compound) receive special status. This approach can be seen to include a
strongly top-down aspect, as the analytical value of a work is seen to inversely relate to how
well it aligns with a specific formal type. In fact, other theorists have explicitly noted the top-
down aspect of Covach’s approach, saying that his form types are “ideals against which actual
songs can be measured” (Stephan-Robinson 2009, 105). It is unclear how the categories of
simple verse, simple verse-chorus, and contrasting verse-chorus act as “ideals” in this regard,
though, since Covach posits no particular ordering of sections for these form types. In fact,
various configurations of these form types can be found in the examples that Covach provides
(2005).
Overall, the categorization of songs into one of these five form types is a central task
in the analytical system that Covach develops for rock music. Many of the historical and
stylistic claims he makes rely on the form category to which a song is assigned (2003). As
well, the artistic merit of a song is seen to derive significantly from the extent that it departs
from a specific form type, particularly those in an AABA pattern (2006).
Form in the writing of Walter Everett
The conception of large-scale form found in the writings of Walter Everett is much
looser than those found in the work of either Covach or Stephenson. Everett, in fact, gives
very few specific form types with which to categorize songs. When these form types are
presented, moreover, they are only briefly described.
Everett explains his point of departure for large-scale form rather succinctly in his
2009 book. He states that songs have “sections that function as verse, chorus, or bridge,” and
that songs “then combine a number of these different sections to create contrast in a balanced
presentation” (141). This generic scheme, according to Everett, acts as a “norm” within rock
music. Like other authors, Everett focuses his approach to form on the section categories of
Chapter 2: Background 31
verse, chorus, and bridge. But he does not prioritize any particular arrangement or
combination of these parts. He offers a bit more detail in an earlier essay, in which he states
that a pop song “typically alternates verses… and choruses… [which will] usually be balanced
by one or two statements of a contrasting bridge” (2008b, 113). Overall, though, Everett does
not seem concerned with grouping songs into different types based on the way these sections
are ordered or used. If anything, we could say that – as his central form type – Everett posits
something akin to Stephenson’s verse-chorus-bridge form.
Everett does mention what he calls the “one-part form,” which consists of a single
part repeated throughout the song without any contrasting material. Like Stephenson,
Everett defines the “material” and “part” through harmonic means. A song that is in a one-
part form, for instance, may contain only a single section type (either a verse or chorus); or, a
one-part song may contain both a verse and chorus (where the verse and chorus together
function as the “part”) (2009, 142). Everett’s one-part form is thus identical to Stephenson’s
one-part strophic form and overlaps with Covach’s simple verse and simple verse-chorus
form types. But unlike these other authors, Everett takes a somewhat pejorative view of this
particular form type. Notably, he states that “the unabating repetition [in the one-part form]
can be a major flaw despite strong material” (2009, 142), after which he provides as an
example the song “Where Did Our Love Go”(The Supremes, 1964). Again, the harmonic basis
for the one-part form type is obvious in this example, as this song can be seen – despite its
one-part status – to contain both verse and chorus sections as well as an instrumental bridge.
One other form type that Everett mentions is the “large SRDC” (2009, 141). (“SRDC”
stands for Statement - Restatement – Departure – Conclusion and will be discussed in much
greater detail in Chapter 4.) The large SRDC is a lot like Covach’s AABA form type; in fact,
Everett states that the AABA structure is a very common instantiation of the large SRDC form
type. Yet the large SRDC accounts for a broader set of songs than would be included in
Covach’s AABA category. For example, a song in large SRDC form may – after the middle “D”
section – include new material or only an abbreviated version of earlier material. (SRDC thus
encompasses both AABA and AABC patterns.) As well, Everett seems to allow for almost any
section type to participate in the large SRDC pattern. The “D” section, for instance, is
discussed as possibly containing chorus or bridge material (143). This view noticeably
departs from those of both Stephenson and Covach, who consider the contrasting middle
material to always be a bridge section.
Aside from these specific cases, Everett does not posit any further form types.
Moreover, form types do not appear to be a central concern within his analytical writing. One
potential reason why form types play a relatively small role in Everett’s analyses is that
Everett brings a strongly Schenkerian outlook to his work. In his 2008 essay “Pitch Down
Chapter 2: Background 32
The Middle,” Everett clearly states that aspects of pitch – especially the tension between a
Schenkerian graph and the surface of a piece – are “of central interest” (139). Thus Everett
can be seen to preference a “bottom-up” approach to form. For Everett – like many
Schenkerians – the “structure” of a song relates more to the organic embellishment of a
simple tonal framework than it does to any foreground arrangement of themes and sections.
Section labels, therefore, are useful in their ability to communicate a song location to a reader
or listener. Otherwise, these section labels appear to offer Everett little analytical grist.
Theories in analysis
In general, there appears to be a fairly high level of agreement between theorists with
regard to basic large-scale form categories. Many of the form types in the work of one
theorist map closely if not exactly to form types in the work of another. Yet certain
differences can be found, and these differences can directly impact our analyses. If we use
Stephenson’s “rounded binary” label, for example, a song exhibiting an AABAA succession
pattern fits the label unproblematically and therefore may not warrant much discussion.
With Covach’s “AABA” label, however, this AABAA pattern becomes the catalyst for an
analytical explanation.
More broadly, we can say that when theorists offer various form types for rock songs,
they are indirectly stating that these form types are meaningfully distinct from one another.
Covach, for example, distinguishes between simple and contrasting verse-chorus forms. In so
doing, Covach implies that it is useful to differentiate between those songs in which the verse
and chorus sections share the same harmonic material and those in which the verse and
chorus sections are built on different harmonic material. Yet it is not entirely clear what
utility this classification scheme holds for analysis. Harmonic factors certainly play a central
role in our perception of form in rock music, but an emphasis of this domain may draw our
attention away from other important domains, such as timbre or melody.
The issue of emphasis also relates to the section roles used in determining large-scale
form types. Because the form types discussed above are based exclusively on the existence
and arrangement of verse, chorus, and bridge sections, these categorization methods
emphasize certain section roles over others. If we use these form types in analysis, we are
consequently encouraged to identify verse, chorus, and bridge sections in order to categorize
a given song. The result is that other section types – such as the link, prechorus, solo, and
refrain – may become subsumed under the verse, chorus, and bridge labels or ignored
altogether. While verse, chorus, and bridge sections are undeniably principle elements in the
form of a song, prioritizing these elements over others can potentially cause us to overlook
some important moments within rock songs (as we will see in Chapter 5).
Chapter 2: Background 33
2.7: Summary
This chapter has laid out a theoretical framework on which the following chapters will
be based. In essence, the epistemological basis of the work herein is derived from research in
the realm of cognitive science. A central premise in this regard is that our method of
conceptualizing categories in general can help us explain categories of musical form in
particular. It seems clear, for example, that a definition-based approach poorly accounts for
our shared understanding of section labels in rock music. Instead, this dissertation employs a
prototype-based approach to form. While the term “prototype” has had a few different
meanings within the realm of music theory (as seen in the work of Brown, Agmon, and
Gjerdingen), the meaning used herein is strongly predicated on a feature-based model, in
which the conceptual category encompasses the attributes of its typical members. With this
model as the axis of a theoretical system, we will see that many interesting aspects of form in
rock music can be revealed.
As a closing thought, it is worth pointing out that – since this dissertation is primarily
concerned with the categorization of song sections – one underlying issue is the identification
of the objects that are to be categorized. As we will see, it is not always clear what exactly
constitutes a “section” in a rock song, and we often find ourselves asking questions such as:
“Should this 16-bar passage be considered a single 16-bar section or two 8-bar sections?”
Note that the issue of what object should be categorized is generally not a concern with other
well-studied domains in cognitive psychology. When we are trying to decide whether a car
seat, for example, belongs to the category of “furniture,” we take for granted that we know
what object we are trying to categorize. Musical form thus involves a two-pronged problem
for categorization, in that we have to not only make category assignments but also decisions
as to what exactly constitutes the object of inquiry itself. This aspect is an important way in
which musical form presents a rather different kind of problem from those discussed in
research on conceptual categories. That being said, this issue comes into play only with those
ambiguous cases (which constitute the bulk of Chapters 4 and 5). Prototypical instances of
section roles (discussed in Chapter 3) – by their nature – offer clear evidence of their status as
discrete sections (or portions thereof).
34
Chapter 3: Roles
3.1: Introduction
Given the context of the prior chapter, we can now turn to using a prototype-based
approach to better understand the categories we use in the analysis of form in rock music. An
initial issue in this regard is which categories to discuss. This chapter covers nine different
section categories – verse, chorus, refrain, bridge, solo, prechorus, intro, outro, and link – all
of which are labels commonly found in the analysis of form in rock music. What kind of
information do these labels convey, though? Sometimes, theorists state that these labels
describe the “function” of a section within the form of the song (e.g., Everett 2009, 141).
Rarely, though, is the function of these section types directly stated beyond the label itself.
Instead, we are meant to infer the function from descriptions, examples, etc. We might
speculate, for example, that a prechorus functions to prepare the chorus; or perhaps the
prechorus functions to transition between the verse and the chorus; or maybe its function is
to provide contrast to the verse and chorus sections. If the prechorus has a single function,
which one of these is it?
Instead of the term “function,” the categories described above will be referred to
herein as section roles. A single role can embody multiple functions. The role of teacher, for
example, includes the functions of instructing, encouraging, disciplining, grading, and so on.
We can thus say that a particular span of music, for instance, fulfills the “role” of (or “acts” as)
bridge material in the song. Knowing that a span of material acts in a particular role tells us
something about how it functions (possibly in multiple ways) within the form of the song.
It should be mentioned that the nine section roles discussed in this chapter do not
represent an exhaustive list of all section roles that have been employed in the analysis of rock
music. Rather, these categories are the most common. In fact, it is not too difficult to find
other labels in use. Covach, for example, uses the label “after-chorus” in his analysis of the
1992 Tori Amos song “Crucify” (2009, 544). Many moments within rock songs defy easy
categorization into one of these nine roles, and authors may prefer to coin a one-time label in
certain instances rather than misapply an existing one. Yet even though various
supplemental labels have been adopted on an ad hoc basis, there is often an effort to reconcile
them with the standard section roles (as evident in Covach’s term “after-chorus”).
Consequently, an understanding of these typical role categories informs our understanding of
those non-typical situations. (This will become especially evident in Chapter 5.)
Chapter 3: Roles 35
The basic structure of the current chapter will be organized around the section roles
themselves. After a brief consideration of a preliminary issue (below), each section label will
be examined in turn. The close relationship between some terms (e.g., verse and chorus) will
require that they be discussed in tandem. In general, the discussion of each role (or roles)
will follow a similar format. First, extant theoretical descriptions of the particular section
role(s) will be summarized. Following this summary, some common usages within analytical
practice will be investigated. This survey of contemporary theory and practice provides a
good foundation for our shared understanding of these section categories. From this
foundation, some prototypical examples of these section roles will then be presented. In
some cases, it will be useful to group these examples into a few subtypes. (Remember, not all
categories can be represented with a single best example.) These prototypical examples will
offer a window into other aspects (i.e., features and attributes) of these section roles that have
not yet been made explicit. Some interesting cases will also be explored to help clarify specific
points.
Measuring a measure
Before beginning this study, one preliminary topic is worth examining first: how
measure lengths are determined. The length of musical spans is undeniably an important
factor in our perception of form in rock music, and length is a central attribute in many of the
form labels that we use. The term “12-bar blues,” for example, specifies an exact number of
measures, and our perception of a 12-bar blues is predicated on this length. Length also
affects our perception of section roles, as we will see, and some section roles seem to have
prototypical lengths. Thus an investigation of section role prototypes involves some
interaction with the issue of length.
Traditionally, lengths of musical units are measured by the measure itself. In other
words, the number of bars in a musical span determines the length of this span. Yet how
much music constitutes a “bar”? There is no obvious answer to this question, since rock
songs do not typically come with a notated score. In general, theorists of rock music have
relied on the rate of rhythmic events to determine measure lengths. Allan Moore, for
example, discusses the “standard rock beat” (2001, 42), which he defines as a snare drum on
the second and fourth beats of a measure. This pattern of rhythmic accentuation is often
referred to as the “backbeat” (Everett 2001, 363), since the even-numbered beats (or the “off
beats”) are more strongly accented than the odd-numbered beats (or the “on beats”).
According to Moore, these off-beat accents alone determine measure lengths. He states
explicitly, in fact, that “the consistent appearance of a snare drum on the second and fourth
beats of a bar allows this length [that of a measure] to be standardized” (42). As the rate of
Chapter 3: Roles 36
off-beat accents speeds up, the length of a measure becomes shorter; conversely, as the rate of
off-beat accents slows down, the length of a measure becomes longer. Measure lengths are
thus directly tied to the pattern of rhythmic accents.
A number of songs, however, provide evidence that determining measure lengths
solely through the backbeat pattern may not be entirely representative of how we perceive
musical spans in rock music. In some cases, the snare drum pattern changes at different
points in the song, even though the measure lengths seem to stay the same. Consider the
song “Should I Stay or Should I Go?” (The Clash, 1982), for example, in which the frequency
of snare hits in the chorus (beginning at 1:08) is twice that found in the verse (beginning at
0:17). One might say that rate of measures in the chorus is double that found in the verse.
Yet the verse and chorus of this song share the same basic harmonic rhythm. In fact, both the
verse and chorus sections could easily be construed as iterations of a 12-bar blues, where the
verse has a normal tempo and the chorus has a “double-time” feel. If the backbeat alone were
to determine measure lengths, though, we would say that the chorus is a 24-bar blues. But
calling the chorus a 24-bar blues when it is basically a variation of the 12-bar-blues verse
seems to misrepresent the form of the song (and the chorus). Tempo, as measured through
rhythmic accents, thus does not seem to always be directly tied to measure lengths. In other
songs, the same section appears multiple times in a song with different backbeat patterns.
Take, for example, “A Day in the Life” (The Beatles, 1967), in which return of the main
material (at 3:18) has twice as many snare hits as its original instance (at 0:56). We could
renotate the reappearance of the section according to this new backbeat, thereby creating
twice as many measures for the same musical material. This renotation would, of course,
capture the perceived change in tempo. But if measures are to be a useful tool for
determining lengths in a theory of form, it seems somewhat problematic to have different
measure lengths for what is basically the same musical span. The issue of measure lengths
versus tempo arises also when we consider that two separate versions of the same song can
have different rhythmic patterns, such as is found with the song “Horses” (by Will Oldham).
The 1997 version (Palace Music) has a snare drum pattern half as fast as the 2004 version (by
Bonnie “Prince” Billy) in terms of the harmonic and melodic pacing. Should measure lengths
in such cases reflect the general harmonic/melodic pacing or the perceived tempo?
Published analyses show, in fact, that scholars do not always use the backbeat as a
consistent metric in their judgments of measure lengths. In some cases, the analyst may
judge that a measure contains only one kick and snare hit. This approach is used by Chris
McDonald in his transcription of the 1991 Nirvana song “In Bloom” (2001, 360). In a similar
fashion, Jocelyn Neal transcribes Sawyer Brown’s 1991 release “The Walk” in a cut time,
where only one kick and one snare hit fill a single measure (2007, 47). In other cases, an
Chapter 3: Roles 37
analyst might decide that a song section contains more than two off-beat accents hits per
measure. For instance, Dan Harrison views what he calls the refrain of “Good Vibrations”
(The Beach Boys, 1966) as eight bars long, despite the fact that the tambourine hits (which
mimic a snare drum pattern) occur four times within his measure (1997, 45). Sometimes, a
theorist seems to be torn between two different interpretations. For example, Timothy
Koozin states in his analysis of the 1994 Sarah McLachlan song “Elsewhere” that the verse
plus transition spans 10 measures (2008, 277), yet he transcribes this same span (in his
Example 10.4) as if it were 20 measures. The inconsistency in Koozin’s designation of
measure lengths may simply be an error. Yet it appears that Koozin was able to hear the
music both ways.
It is worth noting that William Caplin – in his theory of form in the music of Haydn,
Beethoven, and Mozart – has found it useful to differentiate between “notated” and “real”
measures (1998, 35). As Caplin states, the experiential measure – i.e., what the listener
perceives to be the measure – does not necessarily correlate to the measure as notated in the
score. An analogous situation could be said to exist in rock music. In this regard, we might
consider measure lengths as implied by the backbeat to correspond to the “notated” measure.
Similarly, we could say that there exists some “real” measure, which better reflects our
perception of musical spans as distinct from the prevailing tempo. Unfortunately, Caplin
admits that it is impossible to specify exactly what parameters contribute to our perception of
a “real” measure.
Nonetheless, it is clear that factors other than simply the backbeat itself affect our
perception of measure lengths. What might these other factors be? One possibility is
absolute time. For example, both the verse and chorus sections to “Should I Stay or Should I
Go” last about the same length of time, even though their tempos are different. The effect of
absolute time on our perception of meter has received a fair amount of research in the past
century (see London 2004, 27ff). For example, Parncutt notes a “maximal pulse salience” for
periodicities in the range of 600-700 milliseconds. In other words, listeners prefer to hear a
tactus (i.e., the “beat”) at speeds of about 85-100 BPM (1994). This “maximal pulse salience”
may help explain the situation in “Should I Stay or Should I Go,” as the tempo for the verse is
around 112 BPM, whereas the tempo for the chorus (as implied by the backbeat) is twice that
(around 224 BPM). Potentially, our sense of measure lengths is drawn to the middle-level
tactus, and so the 224 BPM section is heard in terms of “real” measure lengths at a rate of 112
BPM. Unfortunately, the use of a mid-tempo tactus as a measure-length device is not a
foolproof method. Take the song “Maybellene” (Chuck Berry, 1955) for instance. Based on
the snare drum pattern, the song has an apparent tempo of about 240 BPM. This rate lies far
outside the window of maximal pulse salience. One solution would be to cut this 240 BPM
Chapter 3: Roles 38
rate in half. The result would be a rate of about 120 BPM. In fact, the 120 BPM rate coincides
nicely with the alternation of bass notes in the guitar part. (Is the backbeat thus the off-beat
accents in the drum part or the off-beat accents in the guitar part?) This 120 BPM rate may
indeed be where we as listeners find it easiest to tap our foot along with the music. But we
cannot necessarily conclude that this 120 BPM rate reflects the speed of a quarter note in a
4/4 meter. Most problematically, the 120 BPM rate (assuming a 4/4 quarter note) would
mean that – instead of opening with a clear 12-bar blues structure – “Maybellene” opens with
a 6-bar blues. This reading rubs directly against inherited notions of measure lengths and
form types in rock music. Of course, we could maintain the 120 BPM rate and posit a meter
of 2/2, which would preserve the 12-bar blues structure. But this solution only recasts the
problem. When, for example, do we posit a 2/2 meter instead of a 4/4 meter?
Ultimately, a tidy formula for determining measure lengths in rock music is not
currently available. Yet all is not lost. The calculation of measure lengths in many if not most
songs is a clear and straightforward process. As a result, there has been a conscientious effort
in this dissertation to avoid choosing musical examples in which measure lengths are
ambiguous or debatable. Unless otherwise noted, all transcriptions imply either a 4/4 or 2/2
meter – the only difference being the rate at which one perceives “the beat” of the song. That
being said, some especially useful examples will be included that raise the issue of measure
lengths, and the reader may disagree with the chosen metric setting. The gist of the argument
will hopefully shine through in these cases.
3.2: Verse and Chorus
Verse and chorus sections are two (arguably the two) basic section types found in
rock songs. Every conception of form in rock music requires that at least one of these two
section roles exists within a song, and most songs are seen to have both. In most extant
theoretical descriptions, these two section types are discussed separately. This separation
occurs in part because some songs are seen to have only one of these sections (typically, a
verse without a chorus). Yet many qualities of verse and chorus sections are described by
theorists in relational terms. That is to say, the attributes of a verse are construed in relation
to a chorus, and vice versa. Because our conception of verse and chorus qualities appears to
have a strong relational component, these two sections will be discussed in tandem here. As
we will see, though, internal (or non-relational) factors can also be found to associate with our
sense of verse or chorus quality. These internal factors can be helpful in making section
designations in unclear situations, such as single-section songs or songs with multiple verse
Chapter 3: Roles 39
or chorus candidates. Yet the difference between relational and non-relational factors is
somewhat a matter of perspective, as will be shown.
Before delving into current descriptions of verse and chorus sections, one small
caveat is necessary. The terms “verse” and “chorus” are often used in the discussion of songs
that lie outside the rock repertory. This situation is most easily found in jazz, where many
popular songs (“standards”) written in the first half of the 20th century are seen to contain
introductory “verse” and 32-bar “chorus” sections. The use of these same form labels in both
the jazz and rock repertoires is no coincidence, and some authors have offered explanations
for this lexical overlap in terms of the evolution of song forms (e.g., Stephenson 2002, 141).
This relationship will be explored somewhat in the next chapter during the discussion of the
AABA organizational scheme, although a full exploration of the historical development of
these labels is beyond the current scope. Needless to say, while some similarity between the
usages is evident, the terms “verse” and “chorus” are considered by most theorists to carry
distinct and different meanings in rock music as opposed to jazz.
Extant descriptions of verse and chorus sections
Theorists have identified a variety of attributes that comprise our perception of verse
and chorus sections. One of the most basic of these is length. Stephenson touches on this
issue, stating that the length of a single repeating section (whether verse or chorus) is
typically eight, twelve, or sixteen bars (2002, 139). Yet as was discussed in the introduction to
this chapter, determining how much musical material constitutes a single measure can itself
be unclear. Consequently, other theorists concern themselves with relative instead of
absolute lengths. Allan Moore, for example, posits that the lengths of verse and chorus
sections are typically in 1:1 or 2:1 relationships, with the most common scenario being a verse
that is twice as long as the chorus (2001, 52). The size of a section may thus be somewhat
contextual. But – at least in Moore’s conception – the lengths of verse and chorus sections
within a single song are not usually wildly dissimilar.
One common method of identifying different sections within a song, of course, is
simply that a certain span of music differs in some meaningful way from the material
surrounding it. This factor – we may call it “contrast” – appears in numerous descriptions of
verse-chorus relationships. Sometimes contrast is presented as an explicit attribute of this
relationship (e.g., Pattison 1991, 55), while other times contrast is implied through
descriptions of how the two sections might differ (e.g., Stephenson 2002, 132). Contrast also
plays an important role in the categorization scheme used by Covach to distinguish certain
types of verse-chorus pairings (2005, 72). Yet while some element of contrast appears to be
an important aspect of verse and chorus relationships, these two sections are seen to often be
Chapter 3: Roles 40
highly similar. Covach, for example, foils his contrasting verse-chorus form with “simple” (or
“non-contrasting”) verse-chorus form, in which the verse and chorus share the same
harmonic material (2005, 73). Covach’s simple verse-chorus form highlights the fact that we
may perceive separate verse and chorus qualities when the attributes of these two sections
significantly overlap. In fact, multiple theorists describe verse-chorus pairings in which both
sections share the same melody and harmony (Covach 2009, 102; Everett 2009, 145). Thus
while some contrast is necessary to distinguish the two sections from each another, verse and
chorus designations may also rely on aspects of their similarity.
Issues of section identity aside, a central theme in theoretical descriptions is that a
chorus section demands our attention much more strongly than does a verse. Covach refers
to this aspect in terms of focal quality, as he states that “in a verse-chorus song… the focus of
the song is squarely on the chorus” (2005, 71). Stephan-Robinson conveys a similar
sentiment when she notes that a chorus has more “energy” than a verse (2009, 94). Likewise,
Everett states that a chorus often has a more “dramatic” harmonic, melodic, or rhythmic
structure than the verse. Others couch the relationship in terms of our ability to recall the
section, saying that the chorus is a more memorable portion of the song in comparison to the
verse (e.g., Harris 2006, 63; Osborn 2010, 87). In all of these conceptions, there is something
more compelling about the chorus and, consequently, something less compelling about the
verse. In general, these explanations attempt to get at the essence of these two section roles,
although how this “focus” or “memorability” is generated in the chorus is never made entirely
clear.
One factor that seems to relate to this aspect of focus and memorability is the domain
of texture, i.e., dynamics, timbre, and instrumentation. In particular, chorus sections are
seen be thicker in texture – with an increase in instrumental resources and general volume –
as compared to verses (e.g., Everett 2001, 49). Various authors trace the etymological
derivation of the word “chorus” itself to the fact that background singers (the “chorus”) were
often added to this section. Again, it is not surprising in light of this attribute that focal
quality is something theorists associate with a chorus section. When one section is louder
than another, it certainly has a much stronger pull on our attention. The term “chorus” also
implies that we as listeners are meant to sing along – a task that requires a certain ease of
memorability.
Another factor potentially related to focus and memorability is the lyric content of
these sections. Often, for example, verses are found to develop a story or exemplify various
aspects of a recurrent theme (e.g., Everett 2001, 49); in contrast, the chorus delivers a “more
general message” (e.g., Burns 2005, 138). Stephan-Robinson also notes a characteristic
perspective shift in many verse-chorus relationships. A verse might use a first-person,
Chapter 3: Roles 41
individual point of view in order to narrate the action, whereas the chorus will switch to a
second-person point of view in order to provide commentary (2009, 240). The presence of
the title (or some variation of it) is also seen to be a common attribute of chorus sections (e.g.,
Endrinal 2008, 69).
One central factor in all descriptions of verse and chorus sections is the large-scale
pattern of lyric repetition. Sometimes, this factor is presented as a defining attribute of verse
and chorus quality. Harris, for example, states that “verses are recurring sections with the
same music, but different text…. [while] choruses, in contrast, are recurring sections with the
same music and text” (2006, 63). More commonly, though, some flexibility is admitted. For
instance, Covach writes that verses “often” feature new lyrics while choruses “tend” to use the
same lyrics upon future iterations (2009, G-3). In general, chorus sections are widely
described as repeating their lyric content on each appearance in the song, as opposed to verse
sections that typically feature new lyrics. It is not entirely surprising, therefore, that the
chorus is described as the more memorable of the two sections. The more something is
repeated and the more general this content is (as in the lyrics to a chorus), the more we are
prone to remember it.
In describing large-scale text repetition patterns, theorists interact with the question
of how verse and chorus sections are distributed in a song. In particular, verse and chorus
sections are seen to be segments of music that reappear more than once in a song.
(Otherwise, we could not judge whether the lyrics to these sections repeat or not on future
iterations.) Because verse and chorus sections repeat within a song, descriptions of their
typical order can sometimes sound like circular reasoning. Moore, for example, defines the
chorus as “the portion of a song that follows a verse,” while the verse is defined as “the
portion of the song that precedes a chorus” (2001, 223-7). Despite this latent circularity, the
general implication – which is supported in the writings of music theorists – is that verse and
chorus sections usually can be seen to group into larger units (or “supersections”), and that
the verse section typically precedes the chorus in these larger units.
This aspect of grouping may, at least in some cases, relate to harmonic factors. It
should be mentioned that harmony has been a central component of verse and chorus
descriptions. For example, Everett states that verse and chorus sections “nearly always”
prolong the tonic (2001, 49-50). In a similar manner, Neal states that tonal closure (i.e.,
ending on tonic) is a common attribute of verse and chorus sections (2007, 45). Yet Neal
states that verse sections can be harmonically open as well. Tonal closure thus appears to be
something that associates more strongly with chorus than verse quality. Other theoretical
explanations, in fact, correlate an emphasis on tonic more strongly with chorus than verse
arrived. The generating agent for this phrase expansion is the deceptive cadence itself, which
denies – and, one might argue, effectively demands – the more final and satisfying closure of
the major-mode tonic. But notice that because of this phrase expansion, our sense of the
structural dominant for this passage has shifted. Although the dominant chord at the end of
the original tail refrain seemed like the most important harmonic moment in the unexpanded
version of this vocal section, the dominant chord at the end of the expansion seems to now
more strongly herald the end of the section as a whole. As a result, the end of this 14-bar
section may be seen as elided by the beginning of a new section. Since the tonic chord at this
moment acts as both the final hypermetrically-strong downbeat of one section and the first
hypermetrically-strong downbeat of another, the vocal melody in this area (using only the text
“I’ll be watchin’ you”) appears to have both head and tail refrain quality. This final vocal
phrase certainly derives from the original tail refrain, but its shape and location are much
more indicative of a head refrain. Note, in fact, that the melody of this final vocal phrase is
basically identical to the head refrain that opens this final main passage. In some instances,
therefore, head and tail refrain quality can merge to create what – although a clear case of
refrain quality – is not clearly one subtype or the other. (Chapter 5 will explore the
implications of this type of blend in greater detail.)
Conclusion
As we can see, our general category of refrain involves at least two subtypes: the tail
refrain and the head refrain. These subtypes involve a number of attributes in the domains of
music and lyrics. Although there is some intersection with regard to the attributes of these
subtypes, it would be problematic to attempt to characterize the general category of refrain
solely via this intersection. Instead, our perception of refrain quality is predicated on (at
least) these two particular configurations.
It was mentioned above that the existence of a refrain may be more strongly
associated with a particular section role. In certain conceptions, for example, refrains are
seen to only exist within a verse. From the examples seen above, however, this does not seem
to be the universal case. Instead, refrain attributes can be found within various section roles.
Nevertheless, the perception of a refrain may help trigger the perception of a particular
section role. This aspect of refrain quality will be explored further in later chapters.
3.4: Bridge and Solo
The term “bridge” is one that has been applied to a great variety of musical situations.
Consequently, it is difficult to come up with a single definition for this term. As we
Chapter 3: Roles 71
investigate these various situations, we find that a few different usages commonly recur. Like
the refrain, our understanding of the term “bridge” can be seen to rely on a few different
subtypes. The bridge label has also been applied on different grouping levels within the form
of a song. A prototype for the term “bridge” should reflect these multiple meanings. As will
be shown below, our understanding of the role of a “solo” section participates in these
multiple meanings. Consequently, a prototype for solo sections will be discussed here as well.
Extant descriptions of bridge sections
Although extant descriptions of bridge sections differ in terms of their details, a few
attributes seem to be consistently associated with bridge quality. In particular, three general
attributes commonly appear in theoretical descriptions: 1) contrast with other sections in the
song, 2) a lack of harmonic (or tonal) closure, and 3) the particular position of the section
within the song as a whole. This last aspect – more so than the others – is contingent on the
nature of other sections in the song and is thus a strongly relational parameter.
The first quality of a bridge – its contrast with other sections in the song – might be
considered a relatively trivial aspect of this section role. Contrast – one might easily point out
– is inherently required to differentiate one section from another. (If two sections did not
contrast with one another via some dimension, there could be no reasonable way of
distinguishing between the two.) Recall, for example, that “contrast” was one aspect that was
seen to differentiate a chorus section from a verse. Yet in discussions of bridge sections,
theorists emphasize especially the element of contrast and imply that contrast itself is a
primary function of this section role. One glossary entry, for example, states that “a bridge
connotes a section that contrasts with the verse and chorus” (Stein 2005, 328). Everett offers
one possible scenario, in which the verse and chorus sections are “diatonically bland” while
the bridge is “far-ranging and chromatic” (2001, 50). In this scheme, the element of contrast
is framed in terms of the harmonic domain. But theorists do not limit the aspect of contrast
to harmony only. In later writings, Everett discusses bridge sections in terms of contrast
within other domains, such as texture and lyrics (2009, 147).
Aside from this general aspect of contrast, however, specific factors within the
domain of harmony are central to many descriptions. One frequent specification is that
bridge sections typically end on a dominant chord (see, for example, Everett 2001, 50; Stein
2005, 328). Moreover, this dominant chord is usually labeled as a “retransitional dominant.”
The implication of the word “retransitional” is that the dominant chord (and thus the bridge
itself) prepares the listener for a return of thematic material (or a key center) that has
previously been heard. Another feature mentioned in extant descriptions is that bridge
sections often begin with a subdominant harmony (e.g., Stephan-Robinson 2009, 161-2). The
Chapter 3: Roles 72
overall harmonic motion of a typical bridge, therefore, appears to be from an opening
subdominant to a closing dominant sonority. It appears, therefore, as if bridge sections
typically neither begin nor end on a tonic harmony, and we might consequently call such
sections “harmonically open.”
In strong part because of this lack of an opening or closing tonic harmony, bridge
sections are considered to be internal parts of a song. The nature of the term “bridge” itself
implies that a bridge section is conceived as something that spans between two other section
types. Consequently, some theorists advise that if a particular section begins or ends a song,
this section should not be considered a bridge (Stephan-Robinson 2009, 98; Stephenson
2002, 137). The exact location of a typical bridge section is not always precisely specified in
theoretical descriptions, though. Moore, for example, states that a bridge frequently occurs
“somewhere between one-half and two-thirds through [the song]” (2001, 223).
Yet in most descriptions, the particular location of the bridge is a crucial feature of
this section role. In fact, some theorists elevate the aspect of location to be the defining
feature of a bridge. Stephenson states, for instance, that “any passage of music heard first
only after the second statement of the chorus can be called a bridge” (2002, 137-8). Similarly,
Covach – in his discussion of form in rock music – presents both AABA and compound AABA
form types, in which the “B” material is the bridge section itself (2005, 74). In these letter
sequence patterns of form, some main musical unit “A” (whether that musical unit is a verse,
chorus, or combination thereof) undergoes a single repetition, after which new material
(whatever that may be, as long as it is “not A”) must be the bridge. The generic nature of
these alphabetic form labels highlights the importance of position over content. The
retransitional nature of the bridge is apparent in these AABA form types (compound or not),
in that the bridge is seen to simply return to the main repeating unit itself. The use of the
AABA form type also implies that the bridge is not typically repeated after its initial
appearance. Some theorists, such as Moore, specifically emphasize the fact that a bridge is
normally heard only once in a song (2001, 223), although others do not hold this view (more
on that below).
The bridge in analytical practice
Given these theoretical descriptions, a few questions come to mind. One issue is what
influence the nature of the main repeating section(s) of the song (the “A” part) might have on
the bridge itself. If the bridge is designed to transition between (or at least contrast from)
iterations of a main formal unit, then one might conjecture that the role of the bridge is linked
to the nature of the main repeating unit itself. Some evidence of this fact becomes clear if we
compare songs in which the main repeating unit (“A”) is relatively short (e.g., eight measures)
Chapter 3: Roles 73
as opposed to long (e.g., thirty-two or more measures). When the “A” unit is short, for
example, the bridge section almost inevitably appears a second time in the song; the result is
an overall AABABA pattern of sections, which Covach refers to as an AABA form with
“abbreviated reprise” (2006). Song forms from the 1950s and 1960s, in fact, commonly
include two iterations of what theorists label as a bridge section (Harris 2006, 64).
Considering that theorists view bridge sections as typically occurring only once in a song,
some evidence of different subtypes of bridges (similar to subtypes of refrains) seems to exist.
If multiple bridge iterations appear in a single song, one might wonder whether or
not the lyrics to these bridge sections repeat as well. Theorists are mostly silent on this issue,
perhaps because the normative situation of bridge sections does not apparently involve
section repetition. But since the external pattern of lyric repetition is such a central criterion
with regard to verse and chorus identity, it is worth examining to what extent real-world
bridges intersect with our perception of these other section types. Stephenson touches on this
topic when he mentions that – if a bridge does appear twice in a song – it “generally carries
the same text” (2002, 137). This bit of information is an important insight, as we will see.
A more central issue in the analysis of bridge sections is how harmonic openness
should be measured. If a section ends with a dominant chord, we might assume that this
section is harmonically open-ended. Neal, in fact, specifically refers to the dominant chord at
the end of a typical bridge section as a half cadence (2007, 45). Yet some final dominant
chords seem more important (or more half-cadential) than others. A common harmonic
feature of 12-bar blues patterns, for example, is a dominant chord in the twelfth bar (the
“turnaround”). It seems unlikely, however, that this final dominant chord significantly affects
how we perceive the 12-bar blues section as a whole. More importantly, a dominant chord in
the twelfth bar of the blues pattern seems to be a subordinate structural element in
comparison to the more obviously cadential motion in the ninth through eleventh bars (e.g.,
the standard blues cadence: V–IV–I; see Chapter 4 for more details on a prototypical blues
harmonic structure). The function of the opening subdominant chord in bridge sections
could be discussed in a similarly problematic manner. The underlying issue – and a
somewhat thorny one in the context of rock harmony – is how we can judge our perception of
“harmonic openness” aside from the rote act of labeling the closing (and/or opening)
harmonies of a section.
The generic quality of contrast is an aspect that allows for even greater interpretive
leeway. Indeed, theorists hold somewhat different criteria for what constitutes contrast – or
at least, the appropriate type of contrast for a bridge section. For example, one common type
of contrast in rock is the use of an instrumental solo somewhere in the middle of the song.
Whether or not this instrumental break may act as a bridge section is an open question in
Chapter 3: Roles 74
analysis and often appears to be decided on a case-by-case basis. This particular issue of the
instrumental break as bridge will explored more below, although parts of this discussion will
be reserved for Chapter 5. For now, note only that even in a case such as an instrumental solo
– where contrast with other sections seems to be rather extreme – it is not clear if we should
consider this an appropriate type of contrast to qualify as bridge material (e.g., see two
different interpretations of instrumental sections in Stephenson 2002 [138] and Everett 2009
[150]).
Classic bridge sections
As mentioned above, there appear to be different subtypes of bridges (as seen in the
case of the refrain) related to the character and length of the surrounding musical material.
Some of the clearest cases of bridge material occur within songs from the 1950s and early
1960s. Because these examples are found so often within this early period of rock history,
these cases will be referred to here as classic bridge sections, which constitute one subtype of
the bridge category.
The song “Handy Man” (Jimmy Jones, 1960) provides an excellent example of a
classic bridge. A variety of factors contribute to the sense that this song contains a bridge
section. One strong element is the location of the bridge with respect to other sections in the
song (see Example 3.4.01). Specifically, the bridge section occurs as the B material in an
AABA pattern. As Example 3.4.01 shows, two iterations of eight-bar A material (Example
3.4.02) are followed by eight bars of B material (Example 3.4.03), after which a third iteration
of A material is presented. The grouping of this AABA pattern into a single unit is
emphasized by the fact that the musical material immediately before and after the AABA core
does not contain vocals with lyrics of any significance. Because the location of the B material
so clearly evokes bridge quality, one common synonym for bridges from this era is the
“middle eight” (Covach 2005, 69-70).
Example 3.4.01: “Handy Man” (Jimmy Jones, 1960); form chart
Start Mm. Pt. Lyrics 0:00 4 intro --- 0:07 8 A “Hey girls, gather round....” 0:20 8 A “I’m not the kind to use....” 0:34 8 B “If your broken hearts....” 0:46 8 A “Here is the main thing....” 1:00 8 solo --- 1:14 8 B “If your broken hearts....” 1:25 8 A “Here is the main thing....” 1:41 8+ outro ---
Chapter 3: Roles 75
Example 3.4.02: “Handy Man” (Jimmy Jones, 1960); A section
Example 3.4.03: “Handy Man” (Jimmy Jones, 1960); B section
It is not only the location of this bridge that makes it sound so prototypical, however;
its musical relationship to the A section is also a strong factor. To begin with, note that the A
section is fully “harmonically closed”: it both begins and ends on tonic, and it has a clear
cadence (using a clear tail refrain) on the downbeat of the seventh bar. In contrast, the bridge
begins and ends off-tonic. More specifically, it begins on the subdominant and ends on the
dominant, which are the standard opening and closing harmonies that theorists mention in
their descriptions of bridge sections. Additionally, the texture in this bridge contrasts
significantly (at least for the era) with the A section. Instead of the mostly half-note values in
the bass line during the A section, the rate of the bass line in the bridge increases to once
every quarter note. The background vocals change as well from their intermittent
punctuations in the A section to the longer, held notes of the B section. The complete stop in
the texture at the dominant chord at the end of the bridge further highlights its separate
identity from the A section that follows. In sum, this bridge includes both the textural and
harmonic factors of contrast that we expect from a prototypical bridge.
Yet we may be more specific with regard to the features of this bridge. The melodic
phrase structure, for example, neatly groups into two-bar units. In this regard, the vocal
phrase organization is contained within the two-bar units created by the underlying
harmonies. These two-bar vocal groupings are common in classic bridge sections. More
important is the specific harmonic progression found here: IV–I–IV–V. This particular
sequence of harmonies is pervasive among bridge sections from the 1950s and early 1960s,
and as a result, these chords act as a prototypical background structure for bridge harmonies
in many songs from this era. Accordingly, we can conceptualize this particular progression in
more abstract terms using Riemannian function labels, as shown in Example 3.4.04. (This
example also includes the prototypical vocal phrase groupings, shown via the dotted lines.)
Example 3.4.04: Generic phrase organization for a classic bridge
The generic S–T–S–D background structure may be realized in a variety of ways.
Certain harmonic realizations, though, are strongly associated with classic bridge sections
from this era. The three most prominent versions (including the literal IV–I–IV–V sequence)
are shown in Example 3.4.05. In these prototypical realizations, the bridge section is shown
to be eight bars long. This length is by far the most common (and fits within the 32-bar AABA
structure). Of course, measure lengths are sometimes difficult to definitely notate (e.g.,
“Donna” by Ritchie Valens [1958] could be construed as either compound duple or compound
quadruple). Other times, the 8-bar standard is obviously scaled upwards to last a full 16 bars
(e.g., “Will You Love Me Tomorrow” by the Shirelles [1960]). As some small evidence of the
pervasiveness of these particular harmonic progressions, Example 3.4.06 lists 32 songs from
this era in which the harmonic material of the bridge conforms closely – if not exactly – to the
abstract S–D–S–T background. (The fact that these sections are, in fact, considered bridges
is attested to by the analyses of various theorists shown in this chart.)
S T S D&
Chapter 3: Roles 77
Example 3.4.05: Prototypical harmonic realizations for a classic bridge
Example 3.4.06: 32 songs with classic bridge sections
Artist Song Yr Analysis Pg Type The Platters The Great Pretender ‘55 Everett 2009 147 a Elvis Presley Tryin' to Get to You ‘56 Everett 2009 147 a The Five Satins In the Still of the Night ‘56 Temperley 2010 -- a Jerry Lee Lewis Great Balls of Fire ‘57 Covach 2009 105 a Ritchie Valens Donna ‘58 Stephenson 2002 140 a Jimmy Jones Handy Man ‘59 Everett 2009 148 a The Coasters Charlie Brown ‘59 Perricone 2000 161 a Elvis Presley Stuck on You ‘60 Stephenson 2002 139 a The Miracles Shop Around ‘60 Stephenson 2002 140 a Bobby Lewis Tossin’ and Turnin’ ‘61 Everett 2009 148 a The Beach Boys Little Deuce Coupe ‘63 -- -- a The Beatles Chains ‘63 Everett 2001 154 a Hank Williams Your Cheatin’ Heart ‘53 -- -- b Elvis Presley Love Me ‘56 Everett 2009 148 b Johnnie Ray You Don’t Owe Me a Thing ‘57 Everett 2009 147 b Bobby Darin Dream Lover ‘59 -- -- b The Coasters That Is Rock and Roll ‘59 -- -- b Bobby Vee Devil or Angel ‘60 Stephenson 2002 140 b Patsy Cline Crazy ‘61 Temperley 2010 -- b Hank Williams Hey Good Lookin’ ‘51 Covach 2009 42 c The Penguins Earth Angel ‘54 Temperley 2010 -- c The Platters One in a Million ‘56 Stephenson 2002 132 c Andy Williams Butterfly ‘57 Everett 2009 148 c Fats Domino I'm Walkin' ‘57 Stephenson 2002 137 c Sam Cooke You Send Me ‘57 Temperley 2010 -- c Everly Brothers All I Have to Do Is Dream ‘58 Covach 2009 106 c The Poni-Tails Born Too Late ‘58 Stephenson 2002 140 c The Marcels Blue Moon ‘61 Everett 2009 148 c The Shirelles Will You Love Me Tomorrow ‘61 Covach 2005 71 c The Beach Boys Surfer Girl ‘63 Covach 2005 71 c Sam Cooke A Change is Gonna Come ‘64 Temperley 2010 -- c Percy Sledge When a Man Loves a Woman ‘66 Temperley 2010 -- c
IV I IV V
IV I V/V V
IV I IV I IV I V/V V
&a)
&b)
&c)
Chapter 3: Roles 78
Returning to “Handy Man,” it is worth noting that the instrumental solo and classic
bridge sections are labeled as separate entities in the form chart of Example 3.4.01. In
labeling these moments differently, an analytical choice has been made. That is to say, this
labeling scheme implies that the solo is not bridge material. This approach is the one often
taken by analysts when labeling parts of songs with an AABA core, in strong part because the
instrumental solo does not fit neatly into the AABA structure that so clearly helps identify a
classic bridge section (see, for example, Covach’s form chart for the 1957 Jerry Lee Lewis hit
“Great Balls of Fire” [2009, 105]). As well, the instrumental solos in songs from this era often
occur over harmonies from the A section (as is the case in this Jimmy Jones song), making a
bridge label for the instrumental solo even more problematic. In general, classic bridge
sections are seen to be distinct from instrumental solos in a variety of ways (although more on
this in a moment).
It is also worth noting that the bridge section to “Handy Man” repeats after this
instrumental solo. Of the 32 songs listed in Example 3.4.06, in fact, 19 (i.e., more than half)
include a repetition of the musical material from the bridge after the initial AABA core
presentation. (When songs do not repeat the musical material from the bridge, this can
usually be seen to fall out of absolute time issues, in that songs with slower tempos do not get
the chance to repeat the B section without becoming too long.) As discussed above, we might
wonder whether the lyrics to these classic bridge sections repeat as well, since the external
pattern of lyric repetition is a central attribute of verse and chorus sections. Of the 19 songs
in Example 3.4.06 that include multiple iterations of the B section musical material, all but
four also repeat the lyrics to the B sections. This piece of data – although admittedly drawn
from a limited selection of songs – shows compelling evidence that lyric repetition is also
associated with classic bridge sections.
In conclusion, our bridge label can be traced back to a set of clear cases from the early
periods of rock. These cases are referred to here as classic bridge structures. These structures
embody all three of the main attributes – location, harmonic openness, and contrast – that
theorists associate with bridge quality. Yet these classic bridges also contain other distinctive
qualities, such as a particular harmonic background and a tendency for multiple iterations
within a song. These attributes thus trigger our sense of not only a classic bridge but also the
more general category of bridge as a whole.
The harmonic quality of classic bridge sections
As is obvious, the S–T–S–D harmonic background that underlies a classic bridge
avoids tonic harmony at both the beginning and end of the section. But we could also more
broadly say that classic bridge sections generally avoid tonic harmony altogether, especially at
Chapter 3: Roles 79
strong or important moments in the hypermeter. Although classic bridge sections typically
do indeed include tonic harmonies, these tonic chords are placed in the weakest possible
location with the hypermetric framework. For example, in both the prototypical progressions
IV–I–IV–V and IV–I–V/V–V, the two bars of tonic harmony are not located at the beginning
of either the first or second 4-bar hypermeasure. The third prototypical progression for
classic bridges (shown in Example 3.4.05c) includes more iterations of tonic harmony, but
each one arrives on a hypermetrically weak beat and is avoided near the end of the
progression; again, tonic harmony is downplayed with regard to its placement within the
section overall.
As we investigate songs from the middle of the 1960s and beyond, the tendency for a
bridge section to avoid tonic harmony – especially in strong hypermetric or cadential
locations – is found to be a prevalent attribute. Some examples from the Beatles help
illustrate this characteristic. Example 3.4.07 shows the bridge section to the song “Ticket to
Ride” (1965). This bridge appears as the B section in a core AABA pattern, after which the B
and A sections immediately repeat (i.e., there is no instrumental solo section in the middle).
The bridge in this song begins on a subdominant harmony and ends on dominant, as we
expect of a classic bridge section, but tonic harmony is conspicuously absent. By lacking any
tonic harmony whatsoever, the B section to “Ticket to Ride” harmonically contrasts even
more strongly from the A sections (which each begin with a long span of tonic and end on
tonic). Furthermore, the bridge can be said to convey an overall feeling of not just harmonic
openness but harmonic instability. A very similar situation can be found in the Beatles song
“I Saw Her Standing There” (1963).
Example 3.4.07: “Ticket to Ride” (The Beatles, 1965); bridge (B section)
Perhaps because of the emphasis on the subdominant near the beginning of many
classic bridge sections, theorists have mentioned that bridges often explore “flat-side keys”
I don't
IV
know why she's rid in'- so high. She ought ta- think twice, she ought ta- do right by me.
V
Be
fore
IV
she gets to say in'- good bye, she ought ta- think twice, she ought ta- do right by me.
instance, the first two vocal phrases begin with non-syncopated rhythms (“You’re on the
phone with your girlfriend”) and end with syncopation (“she’s upset”). In contrast, the
opening vocal phrases of the chorus begin with syncopation (“If you could see that I’m the
one”) and end in a rhythmically-straight manner (“who understands you”). The prechorus
may be viewed as a large-scale move between these two rhythmic organizational strategies.
The first half of the prechorus uses only non-syncopated vocal rhythms (which are
emphasized in the verses due to their hypermetric location), and the second half uses mostly
syncopated vocal rhythms (which are similarly emphasized in the chorus). The last two bars
of the prechorus, in fact, display the same basic syncopation pattern that opens the first two-
bar unit in the chorus. Whether or not one finds this analysis plausible, we can say – at
minimum – that the prechorus seems to internally transition between two different vocal
syncopation strategies.
Of course, one of the strongest attributes that argues for the material in Example
3.5.03 to be a prechorus is its relationship to other sections within the succession pattern of
the song. As Example 3.5.04 shows, the eight-bar prechorus is in its prototypical location –
after the verse material and before the chorus. Yet it is not simply its location that makes this
prechorus example so clear; it is also its length and its relationship to other section lengths
that causes it to seem like its own section. For example, the verse section consists of two
iterations of the eight-bar verse material found in Example 3.5.01 (albeit sometimes with
variations). The prechorus – being eight bars long – aligns with this basic 8-bar grouping
structure, and so we have strong evidence that it is as structurally important as other 8-bar
units. The regularity of the 8-bar unit, moreover, can as strongly determine what is inside a
section as it can determine what is outside. Had the prechorus only been four bars long
(hypothetically speaking), this prechorus would still have seemed like something external to
both the verse and chorus sections, since the internal 8-bar regularity of these sections
effectively seals them off as standalone sections. It is thus both the hypermetric regularity of
the sections surrounding the prechorus as well as the length of the prechorus itself that
contributes to our sense that it requires its own section label.
Example 3.5.04: “You Belong with Me” (Taylor Swift, 2008); form chart
Start Mm. Section Lyrics 0:08 8 “You’re on the phone with....” 0:23 8
Verse “I’m in the room, it’s a typical.... “
0:37 8 Prechorus “But she wears short skirts....” 0:52 8 Chorus “If you could see that I’m the one....”
Chapter 3: Roles 96
As a final factor, it should be mentioned that the lyrics to this prechorus – for the
most part – repeat on the future iteration of this passage. From the third bar forward, the
lyrics in the second prechorus are basically identical to the lyrics from the first prechorus.
Although the first two bars of the second prechorus do not include the same lyrics as its initial
appearance, the basic structure is very similar (1x [0:37]: “She wears short skirts; I wear T-
shirts” 2x [1:40]: “She wears high heels, I wear sneakers”). Consequently, we might say that it
is within the prechorus section that we transition from the non-externally-repeated lyrics of
the verse to the externally-repeated lyrics of the chorus. Yet it is difficult to precisely say how
external patterns of text repetition (or lack thereof) relate to prechorus quality. (My own
sampling of clear prechorus sections shows that – more often than not – its lyrics do, in fact,
repeat on future iterations.) Nonetheless, given that a typical prechorus moves away from a
verse and towards a chorus, it seems plausible that text repetition on future iterations can act
as a cue that the focal moment of the song is about to arrive. (The label of “prechorus”
emphasizes its role as harbinger of the chorus rather than as the conclusion to the verse.) In
other words, by repeating text from earlier iterations, the prechorus may communicate to the
listener that something important is happening or about to happen.
Prechorus as intermediary contrast between verse and chorus
In the discussion of “You Belong with Me,” we were able to see how our perception of
prechorus quality derives (in an abstract way) from aspects of transition, and that transitional
elements may manifest in a variety of domains, such as rhythm, lyric repetition patterns, and
melodic register. In Chapter 5, the implications of the relationship between a sense of
transition and our perception of prechorus quality will be explored further. As we might
expect of a prototype-based approach, of course, prechorus quality is not necessarily
predicated on transitional elements. Indeed, we may also cast prechorus quality simply in
terms of the contrast it provides between clear verse and chorus material. The song “Building
a Mystery” (Sarah McLachlan, 1997) provides a good illustration of this situation. In this
song, the verse (Example 3.5.05) and chorus (Example 3.5.05) sections are quite similar on
numerous levels. (The choice of section role labels here is corroborated in Koozin 2008.) As
seen in “You Belong with Me” above, the verse and chorus sections of this song share the
same harmonic content. But the similarity goes beyond this harmonic relationship. Note, for
instance, that the general range explored by the melodies in the verse and chorus sections are
highly similar, as both stay confined to about a third above and below the tonic pitch. There
also does not seem to be any dramatic difference in texture between subsequent iterations of
verse and chorus. Without intervening material, therefore, a direct succession from verse to
Chapter 3: Roles 97
chorus would potentially create only a weak sense that a new section had arrived. In other
words, the lack of a prechorus would arguably weaken the focal quality of the chorus itself.
Example 3.5.05: “Building a Mystery” (Sarah McLachlan, 1997); verse
Example 3.5.06: “Building a Mystery” (Sarah McLachlan, 1997); chorus
Example 3.5.07: “Building a Mystery” (Sarah McLachlan, 1997); prechorus
But of course, “Building a Mystery” does contain a clear prechorus section, as shown
in Example 3.5.07. This prechorus presents a significant departure from the music of the
verse and chorus sections, especially in terms of harmonic content. For one, the harmonic
pace changes (slowing down now instead of speeding up as in the Taylor Swift example). As
You live
vi
in a church
IV
where
I
you sleep with voo doo-
V
dolls. And you
won't
vi
give up the search
IV
for the ghosts
I
in the halls.
V
&0:48(orig. D)
!
&
œj œb œ œ œ Œ œbJ œ œbJ œ œ œb œ œJ œ œnJ œ œ œ
œ œ œ œ œ œ œJ ‰ Œ Œ œ œ œ ™ œj œ œ ˙
vi
'Cause you're work
IV
in',- build
I
ing- a mys
V
ter- y,-
vi
Hold
IV
in'- on
I
and hold
V
in'- it in.
&1:24(orig. D)
&
Ó œ œ œ ˙ Œ œ ˙ œ œ œ œ œ ˙
œ Œ Œ œ œ œ œ ˙ œ Œ œ œ œJ œ œJ œJ œ ™ œ
II
You're so beau ti- ful,-
IV
with an edge and a charm, and...
II
You're so care ful- when
IV
I'm in your
V
arms.
&1:12(orig. D)
&
Œ ‰ œJ œ œ œ œ œ Ó Ó œ œ œ œ œ ˙ œ
Œ ¿ ˙ œ œ Œ œ Œ œ Œ œb œ ˙ ™
Chapter 3: Roles 98
well, the generic diatonic content of the vi–IV–I–V progression found in the verse and chorus
sections is foiled by the non-diatonic addition of a major II chord. (We might also consider
this to be a Dorian IV chord given a pitch center of B minor for the song overall.) While it
may not be clear whether we should consider the verse and chorus sections in this song to be
tonally stable, certainly the harmonic content of the prechorus is markedly less stable.
Overall, the effect of the prechorus is one of obvious departure from the sound world shared
by the verse and chorus sections. When the chorus begins, consequently, it sounds much
more like an arrival or a homecoming because of the departure generated by the prechorus
itself. The prechorus section can thus be said to strengthen the focal quality of the chorus in
the face of great similarity between the verse and chorus material.
This ability of a prechorus to strengthen the sense of arrival or focal quality of the
chorus is an important aspect of the prechorus role. As a final example, consider the song
“Bad Romance” (Lady Gaga, 2009). The sections of this song are not transcribed here, as the
effect of the prechorus (beginning at 1:05) is predicated primarily on mechanisms other than
melody and harmony. In fact, it is hard to say what the melodic and harmonic content of this
prechorus in this song is exactly, since the voice becomes more spoken than sung and few
supporting pitch-based elements persist. In essence, the bottom drops out of the music as the
texture becomes dramatically reduced. There is thus no clear element of transition between
verse (starting at 0:33) and chorus (starting at 1:13) sections during the prechorus. Instead,
this prechorus creates a high level of contrast (in a variety of domains) between the verse and
chorus material. Unlike the Sarah McLachlan example, however, the verse and chorus
material of “Bad Romance” are not very similar. Nevertheless, we could say that the verse
and chorus sections of “Bad Romance” are more similar to each other than they are to this
prechorus section. Again, the prechorus section generates a stronger impact and sense of
arrival for the chorus section than would have existed had the verse material progressed
directly to the chorus without any intervening material. In this regard, the prechorus
amplifies and clarifies the focal quality of the chorus section that follows.
Conclusion
In summary, clear prechorus quality is conveyed not only through the attributes of
the prechorus section itself but also by the attributes of the sections both before and after it.
Specifically, clear and standalone verse and chorus sections will contribute to the sense that
material found in between them acts as a prechorus. As we have seen, one prototypical
function of a prechorus is to transition between clear verse and chorus sections in a variety of
domains. Yet another prototypical function of the prechorus is to provide contrast between
similar verse and chorus material.
Chapter 3: Roles 99
To a certain extent, the aspects of “transition” and “contrast” potentially present
opposite qualities for our perception of prechorus quality, especially as found in the examples
provided above. On the one hand, the transitional elements of a prechorus cause it to be
similar to verse and chorus material. On the other hand, the quality of contrast can cause a
prechorus to depart greatly from verse and chorus material. This apparent contradiction
creates problems for those attempting to define what exactly constitutes a prechorus section.
Yet it is these contradictory functions that play a strong part in our understanding of the
prechorus role.
In the discussion of the prechorus section, the reader may have noticed many
similarities between this section role and that of a bridge. As we will see, in fact, many
attributes of a clear prechorus section overlap with those of a classic bridge. From the
perspective of harmony, for example, the first supersection of “You Belong with Me” (shown
in Example 3.5.04) can be organized into a basic AABA pattern. Consequently, some
connection between the verse-chorus and AABA forms seems evident. These insights will be
tabled for now until Chapter 4, where a more complete discussion of AABA patterns is
possible.
3.6: Intro, Outro, and Link
The categories of refrain, verse, chorus, bridge, solo, and prechorus stand as a
valuable group of form labels for the analysis of rock songs. Most musical material in rock
songs interacts with or evokes qualities of these section roles in some manner, and many
songs include clear examples of one or more of these roles. That being said, a few remaining
song components are worth discussing. Specifically, the terms “intro,” “outro,” and “link” are
useful to describe song sections that – in their prototypical cases – clearly fall outside the
scope of the refrain, verse, chorus, bridge, and prechorus labels.
It should be mentioned that the terms “intro,” “outro,” and “link” are not necessarily
the standard category labels shared among music theorists. Instead of “intro,” for example,
many theorists prefer the longer term “introduction” (a small difference, admittedly). The
term “link” has also been called “interlude” by some authors (see below). Similarly, the term
“coda” seems to be generally preferred over the term “outro” (e.g., Stephan-Robinson 2009,
102-3). Stephan-Robinson’s explanation of why she prefers the term “coda” is instructive
here and explains my own preference for the terms “intro,” “outro,” and “link.” She writes
that, although the term “outro” is used by musicians and writers, it has a “nonacademic
connotation.” Consequently, she prefers the term “coda” (presumably because it has an
academic connotation).
Chapter 3: Roles 100
Yet it is precisely because the term “coda” has an academic connotation that it will be
avoided here as the basis for a cognitive category. The term “coda” carries with it a long and
rich history of common-practice musical conventions that does not necessarily apply in
modern songwriting. In other words, the term “coda” potentially evokes more exemplars and
prototypes than are valid for rock music. That is not to say that one cannot find attributes of
classical codas in rock songs per se. Rather, an outro and a coda are not necessarily the same
thing. Since the term “outro” is commonly used in the songwriting community and – perhaps
more importantly – not used much outside of rock, the term “outro” serves as a better
cognitive locus for this repertoire. The terms “intro” and “link” are preferred here for similar
reasons.
Intros, outros, and links in theory
If we consider the verse, chorus, refrain, bridge, solo, and prechorus sections to be
the primary parts of a song, then any moments not covered by these terms might be
considered secondary or “spare” parts. This subsidiary status is generally assigned to intro,
outro, and link sections within the hierarchy of a song. Everett, for example, states that intros
and outros are “the least important from a structural view” (2009, 152). Similarly, Stephan-
Robinson says that these section roles are “formally inessential” (2009, 92-3).
Because of the relative unimportance accorded these sections roles, we should not be
surprised that explanations for intros, outros, and links are universally brief. The sections
labels themselves arguably provide sufficient information as to their primary functions: an
intro begins a song, an outro ends a song, and links are the “various musical spacers that may
exist between the main sections” (Harris 2006, 64-5; Harris uses the term “interlude” instead
of “link”). This description adds further evidence that identifying the “main sections” of a
song is a significant factor for intro, outro, and link identity – perhaps a more significant
factor than any internal attributes that these secondary sections may display.
Nevertheless, one internal attribute does seem to be a central factor. Specifically,
theorists note that intros, outros, and links often do not contain a lead vocal part. This aspect
helps explain the secondary status accorded to these sections. The primary parts of a song –
one could easily argue – are those parts that include singing. (That is why we call it a “song”
after all.) Since intro, outro, and link sections are often instrumental, it is understandable
that one might consider them subsidiary to those sections that include vocals and lyrics.
There are reasons to disagree with this stance, but let us table this topic for the moment.
In some conceptions, the instrumental nature of intros, outros, and links appears to
define these section types. For example, Stephenson states that “any instrumental music
occurring before the entrance of the voice is called an introduction” (2002, 134). Similarly,
Chapter 3: Roles 101
Endrinal states that a link is “a short instrumental passage that is used between two major
sections” (2008, 68-9). Yet these authors may not intend to have their descriptions so
narrowly interpreted. Certainly, we can find looser descriptions, in which intros, outros, and
codas are said to only “usually” be instrumental (e.g., Harris 2006, 64).
Further details on internal attributes of these section types are not available. Some
general aspects are worth mentioning, though. In particular, intro, outro, and link sections
are seen to often be based on the same musical material. Moore, for instance, says that
“introductions may recur as playouts or breaks” (2001, 52), by which we should interpret
Moore to be saying that the music to the intro may recur as the music to the outro or link
sections. Additionally, theorists note that this shared musical content often derives from (or
is identical to) musical material found in the verse and/or chorus sections (Harris 2006, 64).
It seems, therefore, as if many intro, outro, and link sections may be similar to or even
indistinguishable from each other or the surrounding material.
Intros, outros, and links in analysis
In general, theoretical descriptions offer an extremely loose framework through
which to understand these section labels. Any musical content or situation appears to
possibly stand as an intro, outro, and link – as long as it occurs in the correct temporal
location (the beginning, end, or middle of the song, respectively) and is not obviously a main
section of the song. Consequently, analysts have great leeway in the application of these
labels to the form of particular songs. This loose framework operates under the assumption,
however, that main sections (and their boundaries) are always easy to identify. Of course,
this is not always the case. We might wonder what effect these unclear situations have on our
perception of intro, outro, and link sections (and vice versa). As we shall see, it turns out that
the distinction between a main section and a subsidiary section is not always straightforward.
It is also not clear what value necessarily exists in creating – out of hand – a
distinction between main and subsidiary sections. Many instrumental sections – such as the
guitar solo – obviously have important structural roles within the form of the song. As well,
many intro, outro, and link sections are undeniably focal moments in a song. A signature
guitar riff, for example, often acts in this capacity. It would be difficult to argue that the
famous riff from “Walk This Way” (Aerosmith, 1975) is not a main section of the song, despite
the fact that no vocals ever occur along with it.
Another issue when dealing with intro, outro, and link sections is how they should be
treated in terms of large-scale form. Especially in the case of link material, analysts have
adopted a variety of approaches as to what relationship these sections have to the other
sections of the song. It will be helpful in this regard to consider analyses of four different
Chapter 3: Roles 102
songs as found in Covach 2009. (Note that Covach uses the term “interlude” instead of link.)
In “Thank You (Falettinme Be Mice Elf Agin)” (Sly and The Family Stone, 1969; Example
3.6.01a), Covach labels the link as a distinct section from the verse and chorus, and this link is
grouped as the final part of a larger supersection (370-1). While Covach also considers the
instrumental link material in “One” (Metallica, 1988; Example 3.6.01b) to be a separate
section, it is now grouped as the opening section in a larger supersection (494-5). Note that
in neither song does the link material consistently appear as “post-chorus” or “pre-verse”
material. After the third chorus in “Thank You (Falettinme Be Mice Elf Agin),” for instance,
no link occurs before the onset of the “contrasting verse” at 3:19. Similarly, the third link in
Metallica’s “One” does not precede another verse but rather comes before a third chorus. In
other analyses, the link material is subsumed within the neighboring verse or chorus sections.
In his form chart of “More Than a Feeling” (Boston, 1976; Example 3.6.01c), for example,
Covach shows that the second verse is fifteen bars long (as opposed to eleven bars like the
first verse) because it contains four bars of instrumental (i.e., link) material at the beginning
(418-9). The opposite approach is taken in the analysis of “All I Wanna Do” (Sheryl Crow,
1994; Example 3.6.01d), where the link material is positioned within the end of the chorus
(530-1). What appears to be a sixteen-bar chorus starting at 1:06 is actually twelve bars of
clear chorus material plus four bars of an instrumental link (Covach himself admits this fact).
Example 3.6.01: Four different treatments of an instrumental interlude (link) a) Interlude as post-chorus: “Thank You (Falettinme Be Mice Elf Agin)” (Sly and The Family Stone, 1969); form chart in Covach 2009 (370-1)
It would be nice to say that the succession pattern shown in Example 3.7.01 is a
typical succession pattern for rock songs. While the abstract AABA pattern is indeed quite
pervasive, the specific organization in Example 3.7.01 is actually rather rare. In fact, it is very
difficult to find songs that contain clear examples of every single section role distributed in a
pattern that constantly and consistently reinforces section identity. As one possible
explanation, we could posit that such songs would simply be too long. In “You Might Think,”
for example, the section lengths are on the shorter side. (N.B. The chorus to this song is
basically an 8-bar unit, but its last measure is metrically reinterpreted in every case to be the
first measure of the following section.) While the succession pattern in this song thus clarifies
section roles, the lengths of the sections might compromise section clarity. (Is the 4-bar span
labeled as “prechorus” long enough to stand as its own section?)
The vast majority of songs, in fact, do not contain clear examples of every single
section role. One song might not include a solo, for example, while another might have no
link material. Consequently, there are almost always situations that challenge our
categorization process, and these situations hold great potential for ambiguities and blends
between section roles. Many of these ambiguities and blends occur again and again in rock
music. This topic requires its own dedicated discussion, though, and cannot be fully
addressed at this point. The reader will thus have to wait until Chapter 5, in which a full
exposition of this topic will take place.
Chapter 3: Roles 113
Compilation of role attributes
As a final summary of this chapter, it is worth taking stock of the section role
attributes that have been discussed herein. In the chapters that follow, the central narrative
is strongly predicated on these section role attributes. The reader could comb through this
preceding pages – role by role – to compile a list of all the role attributes that were presented.
To save the reader this hassle, though, it seems more useful to simply provide such a
compilation in one convenient location. Accordingly, listed below are the nine section role
categories, along with attributes found to be associated with those roles. When appropriate,
these attributes have been organized into specific domains (designated via italics) or role
subtypes (designated via boldface).
Before presenting this compilation of role attributes, however, a strong warning
should be given. Foremost, the reader should remember that our conceptual understanding
of these section roles is more complicated than any explicit listing of attributes can provide.
This chapter has identified some but certainly not all of the aspects that factor into our
perception of these categories. Some of these aspects are relatively noncontroversial, while
others represent a significant element of conjecture. As well, some significant issues remain
unresolved, such as how to weight these attributes and the extent to which these attributes
interact. A list like the one below inherently runs the risk of overemphasizing certain features
by their inclusion or underemphasizing others by their exclusion. One should thus be careful
not to treat the compilation below as simply an expanded set of definitions, as if the lack of
one or some of these attributes were directly correlated to the degree of prototypicality. The
correlation between lack of attributes and prototypicality is certainly true in a general way.
But our perceptual process is more nuanced than a simple one-to-one relationship.
The reader should also notice that the wording of each attribute in this list differs
from that found in a standard definition. Instead of saying, for example, that a prototypical
chorus section “has text repeat on future iterations,” the list below posits that “having text
repeat on future iterations” is an attribute of a prototypical chorus section. (Note the change
from “has” to “having.”) This difference in wording may seem trivial, but substantive reasons
exist to adopt this particular grammatical structure. Take, for instance, two seemingly
paradoxical attributes found with regard to verse sections. In some cases (such as in Covach’s
“simple verse-chorus form”), a clear verse might prolong a tonic harmony. In other cases
(such as in Covach’s “contrasting verse-chorus form”), a clear verse might emphasize tonic
harmony less strongly than the chorus section. We cannot say, therefore, that a prototypical
verse section both prolongs tonic and avoids tonic without sounding contradictory. Yet our
perception of verse quality is influenced by factors of harmony in both cases. The main issue
here is that there is no single prototypical verse situation but rather different (although
Chapter 3: Roles 114
related) situations that can trigger our perception of a verse role. Therefore, “prolonging
tonic harmony” is one attribute that can influence our sense that a span of music acts as a
verse, while “emphasizing tonic harmony less strongly than another main repeating section”
is an attribute that can also participate in engendering verse quality.
Overall, structured information is somewhat difficult to represent in a feature list. A
few subtypes of section role categories are given (along with their associated features), but
most role attributes are presented in an unstructured way. Perhaps it would be worth
developing a few structured verse and chorus subtypes. (Covach’s “simple” and “contrasting”
verse-chorus subtypes might be useful distinctions in that regard.) As of yet, though, no
significant correlations between features have been found to warrant any further partitioning
of these categories.
• CHORUS •
General: acting as a main section role; being a highly memorable part of the song; contrasting
with verse material; having the highest focal quality of song; having a more dramatic musical structure than the verse; having musical material that is basically the same on future iterations; conveying a sense of arrival
Position: recurring at a future point (or points) in song; being 8-16 bars long; being shorter than verse material; following a prechorus section; following a verse section; preceding a link; not presenting the first vocal material of the song
Texture: being the loudest part of the song; having more energy than the verse; including background singers; having a thick texture or instrumentation
Lyrics: having text repeat on future iterations; including a high level of internal text repetition; delivering a more general message; including the title of the song; containing a small amount of lyric content
Melody: having slow rhythmic values in the melody; having short melodic phrases; emphasizing ^1 in the melody; descending to ^1 in the melody; having a high melodic register; having melodic phrases begin on or before hypermetric strong beats; having melodic content more unified with harmonic content; having a melody that is rhythmically simple; avoiding syncopation in lower levels of the metric hierarchy; having a cadential quality to the melody
Harmony: providing tonal closure; having motion towards tonic; prolonging an underlying tonic; sharing the same harmonic material as the verse; including frequent arrivals of tonic harmony; having short harmonic motions; having fast harmonic motions; lacking static harmony; having a standalone harmonic quality; including a cadential quality to the harmony; employing Ionian mode; emphasizing tonic harmony
• VERSE •
General: acting as a main section role; contrasting with chorus material; having low focal
quality; being a relatively unmemorable part of the song; having musical material that is basically the same on future iterations; having a melody and harmony that are not unified (in terms of stepwise resolutions)
Position: recurring at a future point (or points) in the song; being 8-16 bars long; being longer than the apparent chorus; preceding a chorus section; following a link; preceding a prechorus; presenting the first vocal material of the song
Chapter 3: Roles 115
Texture: being a quiet part; having a less thick in texture or instrumentation; not including background singers; having less energy than the chorus
Lyrics: having text that does not repeat on future iteration; including a low level of internal text repetition; developing a story or exemplifying aspects of a theme; including a large amount of lyric content
Melody: having faster rhythmic values in melody; violating traditional non-chord tone resolution; having a low melodic register; syncopating the melody at lower levels of the metric hierarchy; having a rhythmically complex melody; emphasizing ^5 in the melody; having melodic phrases begin on or after hypermetric strong beats
Harmony: prolonging an underlying tonic harmony; emphasizing tonic harmony less strongly than another main repeating section; having a standalone harmonic quality; using motion away from tonic harmony; having long harmonic motions; having infrequent harmonic motion; having static harmony; having less frequent arrivals on tonic harmony than the chorus; sharing the same harmonic material as the chorus; employing a non-Ionian mode; when in Ionian mode, avoiding the tonic
• REFRAIN •
General: including the title of the song; providing tonal closure; acting on a lower grouping
level than the main section roles; having only one or two lines of text; conveying a sense of arrival
Tail Refrain: being 4 bars long; tracing a single melodic phrase that begins on the first downbeat of the final 4-bar hypermeasure in a larger section; having a harmonic progression that moves to tonic on the last strong hyperbeat of the final hypermeasure in a larger section; having the melodic phrase end on the last strong hyperbeat of the final hypermeasure; being at the end of a verse; being at the end of an 8- or 16-bar section; being tonally-closed; having the melody move towards the final tonic scale degree; having the melody stop before the end of the larger section; having a strong cadential quality
Head Refrain: having a melodic phrase that ends on the first downbeat of a larger section; being at the beginning of a chorus section; having its harmony begin on tonic with the first downbeat of a larger section
• BRIDGE •
General: including the highest level of contrast with other sections in the song; occurring
roughly halfway to two-thirds into the song; following the second iteration of main material in the song; leading to a return of main material in the song; having a dramatic textural change from main material; lacking harmonic (or tonal) closure; being harmonically far-ranging and chromatic; being harmonically unstable; being modulatory; avoiding tonic harmony; not opening the song; not ending the song
Classic Bridge: recurring at a future point in the song; occurring in conjunction with relatively short main material (8 or 16 bars); acting as B material in an AABA pattern; occurring after material that ends with a tail refrain; occurring after material that is tonally closed; including background vocals; having a thick texture; being found in songs that lack a clear verse-chorus unit; demanding the return of previous material; having lyrics that repeat on future iterations; including a textural break before the return of main material; ending on a dominant chord; beginning on a subdominant chord; beginning off-tonic; ending off-tonic; positioning tonic harmony in a weak hypermetric location; manifesting an S-T-S-D harmonic progression; having melodic phrases group into two-bar units; acting on the same grouping level as main section roles
Chapter 3: Roles 116
Modern Bridge: not recurring at a future point in the song; preceding the final verse or chorus of the song; having a unique texture within the scope of the song; overlapping the return of main material; including tonal closure; occurring in conjunction with relatively long main material (16 bars +); being found in songs with clear verse and clear chorus sections; not demanding the return of the main material; leading to an abbreviated version of the main material; acting on a larger grouping level than main section roles
• SOLO •
General: being an instrumental section; occurring before or after a bridge section; including a
single instrument playing a prominent melody; including a virtuostic performance; occurring roughly halfway to two-thirds into the core of the song; contrasting greatly with other sections of the song; acting on the same grouping level as main section roles
• PRECHORUS •
General: being 4-8 bars long; contrasting with verse and chorus sections; transitioning
between verse and chorus; following a verse section; preceding a chorus section; having a medium-level thickness in texture and/or instrumentation; having a medium-level loudness; recurring at a future point (or points) in the song; having musical material that is basically the same on future iterations; heightening anticipation for the chorus section; strengthening the sense of arrival for chorus section; acting on the same grouping level as main section roles
Lyrics: having a moderate level of internal text repetition; having lyrics repeat on future iterations
Melody and Harmony: having its melodic register be in the middle of the vocal range; being a harmonically unstable passage; beginning off-tonic; ending off-tonic; extending a predominant-to-dominant progression
• INTRO •
General: being an instrumental section; being subsidiary to the main section roles; containing
material from main section roles; containing link material; being the first musical material heard in a song; fading in; acting on a larger grouping level than main section roles
• OUTRO •
General: being an instrumental section; being subsidiary to the main section roles; containing
material from the main section roles (especially the chorus); being the last musical material heard in a song; fading out; acting on a larger grouping level than main section roles
• LINK •
General: being an instrumental section; being 4-8 bars long; being subsidiary to the main
section roles; occurring after a chorus section; occurring before a verse section; overlapping with the end of a chorus; providing cadential arrival for a chorus section; containing the main riff of the song; appearing as part of intro and/or outro material; sounding like both the beginning and ending of a larger grouping; acting on the same grouping level as main section roles
117
Chapter 4: Conversions
4.1: Introduction
The previous chapter presented prototypes for a number of different categories into
which we can assign portions of a rock song. Yet these are not the only categories that we
might use to describe the parts of a song. We might say, for example, that some passage in a
rock song is an example of a 12-bar blues. This label of “12-bar blues” is, undeniably, a
category into which we have assigned the passage. At the same time, we might say that this
same 12-bar blues passage acts as verse material within the song. Note that we have no
problem saying a span of music belongs both to the 12-bar blues category as well as to the
verse category. There are thus multiple ways of categorizing a segment of music within a rock
song, and these different categories are not mutually exclusive. Although this situation might
imply that parts of a song can fall into any and all categories (including multiple categories at
once), we find that such is not the case in practice. Instead, some section labels are used by
analysts in a mutually-exclusive way, while others are not.
Categorization systems
The section categories described in the last chapter – verse, chorus, refrain, bridge,
solo, prechorus, intro, outro, and link – can be seen as belonging to one cohesive labeling
system. In general, categories within the system of section roles are considered to be
mutually exclusive. If we label something as a verse, for example, we are traditionally
positing that this span of music is not, for instance, a chorus or a bridge. This is not to imply
that there is no ambiguity between the categories within this system of section roles. Rather,
the more a section is in one role, the less we consider it to be in another. The fact that section
roles can act on different grouping levels of a song might seem to contradict this aspect of
mutual exclusivity. For example, we have no problem considering part of a song to be both a
refrain and the end of a verse. We can simply say, though, that these section roles are
mutually exclusive within a single level of the grouping structure.
Other labels for song sections can be seen as belonging to a different categorization
system. The category of “12-bar blues,” for instance, is not used by theorists to describe the
role of a section. Rather, to label a section as a 12-bar blues makes no direct statement about
how it functions in the form. That being said, we might have some good ideas about what
types of roles a 12-bar blues plays within a song. Yet the label of “12-bar blues” does not
inherently tell us anything about the role of the musical span within the form of the song per
Chapter 4: Conversions 118
se. (Theorists sometimes present statements to the contrary, but we will see that such
statements are problematic.) As a result, categories such as the “12-bar blues” will be referred
to here as organizational schemes. There are a number of different organizational schemes
that can be found within rock music, including the SRDC and AABA patterns. This chapter
will describe specific instantiations of these organizational schemes in detail. As categories,
these organizational schemes are not always as well-established as the categories of section
roles. The SRDC pattern, for instance, is not a category commonly discussed outside the
realm of music theory.
Unlike section roles, organizational schemes are perceived more by their internal
characteristics than by their relationships with other spans of music. We understand a 12-
blues pattern, for example, primarily through its particular melodic and harmonic structure,
and our perception of a blues section is not dependent on other material in the song. As these
organizational schemes become larger and larger – such as the 32-bar AABA pattern – the
“internal characteristics” of the scheme may encompass a broad chunk of music. In essence,
organizational schemes are categories just like section roles, so our understanding of these
schemes is inherently a complicated matter that involves multiple domains.
The distinction between a section role and an organizational scheme may be seen as
similar to William Caplin’s distinction between a formal function and a formal type,
respectively (see Bergé 2009, 21ff). Caplin states, for example, that “the various ‘formal’
functions are all manifestations of general ‘temporal’ functions. But the formal ‘types’ have
no such determinate temporal expression” (32). A musical “sentence,” therefore, is an
example of a small-scale formal type, and large ternary form is an example of a large-scale
formal type. In contrast, the main theme of a large ternary form is considered to be a formal
function. Thus a musical sentence may function as the main theme of a large ternary form.
In a similar way, a 12-bar blues – as an organizational scheme – may act in a verse role. To
say that function and role – or type and scheme – are equivalent, however, might be too
strong a statement given the great difference in musical style between rock and common-
practice music.
Because our perception of organizational schemes is less contextual than that for
section roles, we can specify some common melodic and harmonic structures that associate
with particular organizational schemes. To do so, we require a more detailed method than is
currently available of talking about structures within the domain of melody. Consequently, a
brief discussion of some typical melodic structures and their relationships within rock music
is warranted.
Chapter 4: Conversions 119
Melodic organization
The examination of melodic organization in rock has received some attention in
earlier work, most notably that of Ken Stephenson (2002). Stephenson refers to a variety of
what he calls “phrase rhythms” (7), a term that embraces both grouping and metric structure.
In this regard, Stephenson makes reference to the work of William Rothstein. It is worth
clarifying the meaning of this term in the work of both authors, for they are not entirely
identical. In his 1989 book, Rothstein describes phrase rhythm as the combination of phrase
structure and metric structure. For Rothstein, phrase structure involves the entire musical
content of a passage, i.e., its melodic, harmonic, and rhythmic organization (13). Yet for
Stephenson, phrase rhythm refers only to the interaction of the melody (i.e., the vocal phrase)
and the metrical structure (7). In other words, harmony does not factor into phrase rhythm
for Stephenson (except, perhaps, in that harmony affects our perception of metrical
structure). In essence, Stephenson is talking only about the phrase rhythm of the melody (the
vocal phrase), not phrase rhythm in general. Accordingly, the term “melodic phrase rhythm”
will be employed here to refer to what Stephenson calls “phrase rhythm.” Melodic phrase
rhythm describes how different vocal phrases lie within the underlying metric framework.
In rock music, the underlying metric framework typically consists of a regular pattern
of 4-bar units, i.e., hypermeasures. As a result, we can sketch out some typical melodic
phrase rhythms for rock music. Stephenson offers a number of different labels for these
melodic phrase rhythms, such as the “2+2 model,” the “extension-overlap model,” and the
“first-downbeat model” (2002, 7ff). These labels will not be used in this dissertation, for
reasons that will become clear in a moment. Instead, the actual phrase lengths within the
hypermetric structure will be directly illustrated, as shown in Example 4.1.01 below. Example
4.1.01 gives a few common melodic phrase rhythms found in rock music, although this
collection is certainly not exhaustive. For reference, Stephenson’s labels are shown here as
well. As should be evident, Stephenson’s “2+2” model – which is named for its two bars of
melodic content followed by two bars of rest – does not actually divide cleanly into two
halves. Rather, the melodic content spans through the first two bars and ends on the
downbeat of the third measure. Many of these melodic phrase rhythms, in fact, can be
characterized more simply by the location at which the initial melodic phrase ends. The first
three melodic phrase rhythms shown in Example 4.1.01, for instance, end earlier and earlier
within the hypermetric framework, while d) ends on the first downbeat of an entirely new
hypermeasure. The last melodic phrase rhythm, e), captures the situation where a long up-
beat ends on the downbeat of the hypermeasure. We could also easily imagine additional
melodic phrase rhythms. Not shown, for example, is one that spans from the first measure all
the way until the downbeat of the fourth measure. Stephenson refers to this melodic phrase
Chapter 4: Conversions 120
rhythm as the “traditional model” (6), which he posits is more rare in rock music than in folk
songs, nursery songs, and 19th-century popular songs.
Example 4.1.01: Phrase rhythms in rock music, à la Stephenson 2002
Of course, these melodic phrase rhythms are highly predicated on the ability to
identify what constitutes a measure of music. With a different metrical reading, we would
interpret the melodic phrase rhythm in a different way. Take, for example, some hypothetical
passage that would conform to the 1+1 model. If we were to read the same span of music with
measure lengths half as long (for a total of eight bars instead of four), then we would see two
instances of the 2+2 model instead of one instance of the 1+1 model. Some of the effects of
measure lengths discussed at the beginning of Chapter 3, therefore, impact our use of melodic
phrase rhythms. Stephenson tries to avoid this issue by framing these melodic phrase
rhythms in terms of the relationship between melodic motion and rest (i.e., 2+2).
Nevertheless, he cannot escape the issue of measure lengths, as evident in the ease with which
we can shift between the 2+2 and 1+1 models.
&a) 2+2 model
&b) (untitled)
&c) 1+1 model
&d) extension-overlap model
&e) first-downbeat model
Œ Ó !
Œ Œ
Œ Ó Œ Ó
Œ
Œ Ó Œ Œ Ó Œ
Chapter 4: Conversions 121
An important aspect of organizational schemes is how these melodic phrase rhythms
relate to one another. To illustrate these relationships, it is helpful to employ a few notational
devices. In Example 4.1.02, three conventions are shown that each represent different
melodic relationships. (Some of these should be familiar from examples in Chapter 3.) In a),
the curved dashed phrase marking shows that the two smaller phrases on the surface of the
music can be conceived as one instance of the larger grouping. In b), the curved dotted lines
represent an abstract grouping – a partitioning of a span of music – that may be instantiated
by various types of melodic phrase rhythms. The abstract grouping shown in b), for example,
could be realized by both Examples 4.1.01b and 4.1.01c. Finally, the dashed bracket in
Example 4.1.02c reflects a parallelism between the two phrases. This parallelism is
sometimes strong, as in the case of a literal repeat. More often, the second vocal phrase is
some variation of the first. A musical parallelism is often reinforced in the domain of lyrics
through a rhyme.
Example 4.1.02: Melodic phrase relationships
Conversion
In the previous chapter, section roles were seen to encompass a number of different
attributes in a number of different domains. When we attempt to reconcile a particular span
of music with sections roles (whether consciously or unconsciously), at least part of the
process involves reconciling the attributes of the musical span with those associated with the
section role. It is not too difficult to imagine, therefore, that – given a certain musical passage
– various attribute changes can shift our perception of this passage from one role to another.
&a) larger grouping
&b) abstract grouping
&c) parallelism
Œ Œ
Œ Œ
Chapter 4: Conversions 122
A musical passage with a low level of internal text repetition, for example, would seem less
verse-like and more chorus-like if its lyrics were altered to create a high level of internal text
repetition. In such a case, we might say that this change transforms the passage from less of
a verse role to more that of a chorus. The term “transformation” has developed a particular
meaning in music theory within the realm of mathematically-based work (e.g., that of David
Lewin, Richard Cohn, and Brian Hyer [see Cohn 1998]). Accordingly, this term (and its
various forms) will be avoided in the context of this discussion.
Instead, the term “conversion” will be used here to describe shifts in our perception
of section roles. The word “conversion” has developed a specific meaning within the field of
linguistics, and it is this meaning that is alluded to here (in a metaphorical way). In
linguistics, conversion describes the process by which a word changes from one part of speech
to another (Denham and Lobeck 2009, 197). For example, the noun “mother” is converted to
a verb when we say, “She mothers him too much.” Such linguistic conversions can occur
between a number of categories. Nouns can be converted to verbs (“the trash” becomes “to
trash”), verbs can be converted to nouns (“to commute” becomes “the commute”), adjectives
may be converted to verbs (“clean” becomes “to clean”), adjectives can be converted to nouns
(“crazy” becomes “a crazy”), and so on. In these cases, the basic word stays the same. Via a
change in context, though, the word shifts from one lexical category to another.
As this chapter will show, similar shifts can be seen within rock music. Often, the
same organizational scheme acts in different contexts to convey different section roles. The B
section in an AABA form, for example, sometimes behaves as a bridge and other times as a
chorus. This chapter will primarily explore the conversions of three basic organizational
schemes common to rock music: the 12-bar blues, the 16-bar SRDC, and the 32-bar AABA.
All three of these schemes interact with section roles in numerous ways. In some cases, the
conversion results from a fundamental change to the musical structure itself. For example,
we will see the effects on categorization when a 12-bar blues pattern changes into a 16-bar
blues. In this regard, the use of the term “conversion” departs somewhat from its particular
meaning within the field of linguistics. (The term “derivation” may be somewhat more
appropriate in this case.) Nevertheless, the central process under investigation here is how
similar harmonic and melodic structures can be altered so as to affect significant shifts in our
perception of what role these structures play within the form of a song.
The study of conversion in rock music offers further evidence of the many prototype-
like aspects of section roles. In our investigation of conversion, we can more easily see the
continuum on which these section roles lie. As we move from the discussion of blues to SRDC
and AABA schemes, we discover just how permeable the boundaries between these section
roles are.
Chapter 4: Conversions 123
As a final note, it should be mentioned that almost all of the musical examples in this
chapter are drawn from the 1950s and 1960s, with no examples from the 1980s or later. One
reason for this narrow historical window is simply that clear blues, SRDC, and AABA
organizational schemes seem to be more prevalent in these early years than in later decades.
More importantly, full-fledged verse-chorus-bridge form (to use Stephenson’s term) or
compound AABA form (to use Covach’s term) was not as pervasive during this early era as it
is today (see Covach 2005, 75). Consequently, many songs around the 1960s appear to be
waypoints between older and newer approaches to song form. We could even say that –
within the music of these early years – we see the evolution (or “genesis” as Summach 2011
puts it) toward modern song forms. That being said, the following discussion is not meant to
represent any particular historical narrative of form within rock music. Rather, this chapter
shows how changes to a few particular organizational schemes can affect their relationship
with section roles. Nevertheless, we can see via these conversions the various paths by which
form types may have developed over the course of rock history. The study of conversions thus
offers a potential window into the development of different form types in rock music.
4.2: The Blues
Blues-based harmonic patterns are some of the most common harmonic patterns
found within rock music, especially during the early years of rock. Interestingly, blues
patterns can be found to act in a number of different roles. Sometimes, a blues progression
underpins a verse section. In other cases, it underpins a chorus. We also find situations
where both the verse and chorus sections of a single song are built over blues-based
harmonies. There are also many instances where – despite a clear blues pattern – it is
difficult to judge what section role label (or labels) might be most appropriate for a blues
passage.
Because blues patterns act in these different capacities, they provide a window into
the ways our perception shifts from one section role to another. Within the rock repertoire, of
course, we find a number of different blues patterns – such as the 12-bar and 16-bar blues –
and each of these blues patterns has its own way (or ways) of interacting with section roles.
In fact, it is possible to see – via changes in various domains – how a given blues structure
can be converted from the role of one section to another. One important waypoint in this
process is the “hybrid blues,” which is a particular phrase organization that presents
ambiguous information with regard to section roles. By recognizing the ambiguous nature of
the hybrid blues (and blues patterns in general), we can better understand some of the
persistent issues we face when analyzing rock songs.
Chapter 4: Conversions 124
Classic 12-bar blues structures
Of all blues patterns, the 12-bar blues is by far the one most frequently described in
the writings of music theorists. A central aspect of the 12-bar blues is its pattern of
harmonies, yet exactly what chords make up this harmonic progression is not entirely clear.
In Example 4.2.01, for instance, we see five different harmonic patterns for the 12-bar blues
as proposed by five different authors. While there seems to be general agreement among
these authors, it appears that the category of the 12-bar blues – like other conceptual
categories – defies any strict definition.
Example 4.2.01: Harmonic realizations of the 12-bar blues in various authors
Bar # Source 1 2 3 4 5 6 7 8 9 10 11 12 Everett 2001,54 I I I I IV IV I I V7 V7 I I / V Middleton 1990, 48 I I I I IV IV I I V (IV) I I Moore 2001, 59 I I I I IV IV I I V IV I I Perricone 2000, 154 I I I7 I7 IV7 IV7 I7 I7 V7 IV7 I (V7) Covach 2009, 99 I (IV) I I IV IV I I V (IV) I (V)
In their descriptions of the 12-bar blues, theorists mention not only a characteristic
harmonic structure but also a specific phrase organization. In particular, theorists usually
refer to the 12-bar blues as a bar form (e.g., Everett 2009, 138-9). This bar form may be
represented via a sequence of letters, such as a–a’–b (Everett 2001, 54) or a–a–b (Stephan-
Robinson 2009, 24-25), where each letter refers to a separate 4-bar segment. Alternatively,
the bar form may be conceptualized as a “question-question-answer” metaphor, in which the
“first phrase poses a question, the second poses the same question in a slightly different way,
and the third phrase answers the question” (Covach 2009, 98). These descriptions represent
standard presentations of the 12-bar blues form, and many song examples can be found that
adhere to this format. Consequently, characterizations such as these will be considered here
as referring to a classic 12-bar blues.
As a prototypical example of classic 12-bar blues form, consider the song
“Crossroads” (Cream, 1968), the first vocal section of which is shown below in Example
4.2.02. The general harmonic structure of the 12-bar blues is obvious, and the bar form
should be conspicuous as well. Note, for example, that the second phrase basically repeats
the same text and melody of the first (despite a different harmonic setting), while the third
phrase departs from this rhyme scheme (in concert with the harmonic cadence). It is worth
clarifying what is meant by the term “phrase” here, since theorists often use the term “phrase”
in slightly different ways when discussing the 12-bar blues. Take, for instance, Walter
Chapter 4: Conversions 125
Everett’s statement that the 12-bar blues consists of “three four-bar phrases, with the first two
phrases having the same or a rhyming text and the third a contrasting text” (2001, 54).
Notice how Everett’s use of the term “phrase” involves two separate elements: 1) hypermetric
structure (“four-bar phrases”) and 2) melodic organization (“phrases having the same or a
rhyming text”). As will be shown, it is useful to distinguish between the hypermetric structure
of the 12-bar blues, its harmonic structure, and its melodic phrase structure, since all of these
elements impact our perception of the role this structure plays within a song.
Example 4.2.02: “Crossroads” (Cream, 1968); opening vocal material
Example 4.2.03: Classic 12-bar blues phrase structure
An abstraction of the harmonic and melodic phrase structure for “Crossroads” is
shown above in Example 4.2.03. This abstraction represents a prototypical organization for a
classic 12-bar blues. As can be seen, a classic 12-bar blues has three vocal phrases, each of
which (roughly speaking) begins in a hypermetrically-strong measure, spans two full bars,
and then ends on the downbeat of the third bar. The second long vocal phrase parallels (or
repeats) the first, and each of the large vocal phrases often subdivides into two smaller
subphrases. This prototype also captures some of the common harmonic options of a classic
norm. Yet in very few of these cases were section roles entirely clear. The simplicity of
whatever final section labels we might choose, therefore, hides just how malleable and
tenuous our sense of section quality may be in these situations.
Examples 4.3.24: Relationships between SRDC schemes and section roles
Verse
Refrain
RS D C
b)Verse
Refrain
RS D C C
c)
Verse
Refrain
RS D C
a)
Verse Chorus
RS D C
d)Verse Chorus
RS D C C
e)
Verse Chorus
RS D (new)
(new)
f)
Verse Prechorus Chorus
RS D
g)
Chapter 4: Conversions 178
As with blues-based organizational schemes, we can create a chart of the interactions
between section roles and SRDC organizational schemes, as shown in Example 4.3.24. Once
again, different configurations are bounded by a dotted box, and the simplest path between
configurations is shown via the dotted lines. The chart begins in Example 4.3.24a with a
classic 16-bar SRDC structure (e.g, “I’ll Cry Instead), in which a tail refrain ends what is
wholly verse material. In b), we see this refrain-like area grow (as in “Please Please Me”),
such that – through changes in a number of domains – we reach the situation in d) where the
16-bar SRDC scheme splits into two 8-bar spans of verse and chorus material (e.g., “Ticket to
Ride” or “Drive My Car”).
An alternative path from the classic 16-bar classic SRDC scheme is shown in the
right-hand branch of Example 4.3.24. In c), a doubling of the tail refrain elongates the 16-bar
scheme into a 20-bar structure (as seen in “La-La (Means I Love You)”). In e), this doubled
tail refrain converts into a single chorus section (e.g., “I Heard It Through the Grapevine”).
Through hypermetric reinterpretation, we move beyond any clear sense that there is a full
SRDC structure. This situation is shown in f), where new chorus material supplants the final
conclusion gesture (e.g., “I Can See For Miles”). As a result of this change, we are not far from
a relatively straightforward verse-prechorus-chorus structure, as shown in g), such as that
found in “You’ve Lost That Lovin’ Feelin’.” Again, the reader is reminded that this chart of
relationships between section roles and SRDC structures does not necessarily represent a
historical document of how form types evolved from the early years of rock. Rather, this chart
shows the most direct connections between different configurations. Nevertheless, these
types of conversions certainly play a part in the changing landscape of form types found in the
history of rock music.
4.4: AABA Form
In the previous portion of this chapter, conversions of SRDC patterns were discussed
primarily in terms of a 16-bar framework. This 16-bar framework provided a useful reference
point from which we could compare and contrast different song examples in order to view
shifts in our perception of section qualities. As was mentioned, though, we can find SRDC
patterns in other measure lengths. One example given was the 8-bar SRDC found in the
Beatles song “From Me to You” (1963). We can also find larger groupings of measures, such
as 32- and 64-bar frameworks, that seem to exhibit SRDC construction. There appear to be
some important differences between these longer instantiations of the SRDC pattern and
their shorter brethren, though, as Everett is careful to distinguish between “large” and “small”
SRDC structures (2009, 141).
Chapter 4: Conversions 179
As noted earlier in Chapter 2, one very common instantiation of these “large SRDC
structures” is the form type that some theorists refer to as “AABA.” (This situation occurs
when the conclusion gesture repeats the statement and restatement gestures.) While AABA
forms are similar to SRDC patterns, the ways these AABA forms interact with section labels
differ significantly. One reason for this difference is simply that the constituent parts of an
AABA form are longer than those in a 16-bar SRDC pattern. A 32-bar AABA form, for
example, divides into four 8-bar parts, and these 8-bar parts are much easier to perceive as
standalone sections than the 4-bar gestures in a 16-bar SRDC pattern. As discussed in the
introduction to Chapter 3, however, measure lengths are not always easy to determine. In
some cases, it may not be clear whether we should consider a span of music to be a 16-bar or
32-bar unit. As a result, it is sometimes unclear as to whether part of a song should be
considered a 16-bar SRDC structure or a 32-bar AABA form. Some of the songs discussed
below potentially show the effects of this issue.
Nevertheless, the majority of AABA forms are clearly distinct from their shorter, 16-
bar counterparts. Many AABA-patterned songs, in fact, display characteristic organizational
schemes in the domains of harmony and melody. Like the 12-bar blues and 16-bar SRDC
structures, certain configurations of AABA form are particularly common. Moreover, these
similar configurations act as valuable reference points from which to track shifts in our
perception of section qualities. AABA form thus provides further evidence of the permeable
and continuous boundary between section roles.
AABA overview
As discussed at the end of Chapter 2, theorists present different views on which
section roles may be contained within a standard AABA form (whether this form is called
large SRDC, AABA, or rounded binary). Covach, for instance, consistently views A sections as
verse material and B sections as bridge material (2005, 69). In contrast, Everett views A
sections solely as verse material but allows for B sections to be either a bridge or chorus
(2009, 143). Stephenson presents yet another perspective, as he allows for A sections to
contain either verse or chorus material but refers to B sections only as bridges (2002, 140).
Covach offers the common-ground view here, as all theorists agree that A sections may act as
verse material and that B sections may act as bridge material. Given the descriptions in
Everett and Stephenson, though, A and B sections both seem to potentially act as chorus
material as well.
Although not necessarily explicit, these three views share one important aspect.
Specifically, these descriptions all imply that each letter in an AABA form represents only a
single section role. (Everett and Stephenson present their different labels for the same part
Chapter 4: Conversions 180
using the conjunction “or.”) That being said, theorists do recognize that combinations of
multiple section roles often group into large-scale AABA patterns. Covach’s “compound
AABA” form, for example, describes instances where each A section contains both a verse and
a chorus (2005, 74). Analyses by a variety of other theorists also show multiple section roles
being grouped into large-scale AABA succession patterns (as found in Endrinal 2008,
Stephan-Robinson 2009, and others). It seems worthwhile, therefore, to differentiate
between those songs in which the AABA pattern contains only a single section role per letter
(a “simple” AABA) versus those in which the letters represent multiple section roles (a
“compound” AABA).
The difference between simple and compound AABA forms is not always clear, of
course. Covach’s “incipient” form, for example, describes those cases in which we may be
unsure as to whether the A section contains one or two section roles (2010, 6-7). One central
factor here is the overall size and length of the AABA structure itself. As the A section
becomes longer, there is inherently a greater potential for multiple sections to be contained
within it. In the previous discussion of SRDC patterns, for example, we looked at a number of
songs (e.g., “I’ll Cry Instead,” “Please Please Me,” and “Ticket to Ride”) that could be
considered examples of AABA form. In these songs, a 16-bar SRDC structure served as the A
material, and this 16-bar structure seemed to sometimes split into multiple section roles. 16-
bar SRDC patterns, in fact, are quite common structures for the A material of AABA-
structured songs, and thus we should have a relatively good understanding – given the
preceding portion of this chapter – of some possibilities for the conversion of AABA forms.
Evidence of the continuum between simple AABA form and compound AABA form – at least
from one approach – should thus be quite clear.
Nevertheless, it is worth looking at one AABA example in which the A material
contains a 16-bar SRDC structure, if only because the AABA form as a whole has not been
directly addressed here as of yet. The song “Will You Love Me Tomorrow” (The Shirelles,
1960) presents a typical case. The song includes two main 16-bar units, the first of which
(Example 4.4.01) is clearly an SRDC structure and the second of which (Example 4.4.02)
manifests very bridge-like qualities. These two parts (A and B) organize into a
straightforward 64-bar AABA pattern, as shown in Example 4.4.03. After this 64-bar AABA
pattern, we hear a repeat of the final SRDC structure, although the first half (the SR gestures)
is now an instrumental solo. Repeated iterations of the conclusion gesture serve as the
fadeout material for the song.
Chapter 4: Conversions 181
Example 4.4.01: “Will You Love Me Tomorrow” (The Shirelles, 1960); A section
Example 4.4.02: “Will You Love Me Tomorrow” (The Shirelles, 1960); B section
I
To night- your mine com plete
IV
- ly.-
V
I
You give your love so sweet
V
ly.- To night-
V/vi
the light of love
vi
is in your eyes,
IV
but will you love
V
me to mor
I
- row?-
&(orig. C)0:07
!
&
&
& !
Œ œ œ œ œ œ œJ œ ™ Œ œ ˙ œ Œ
Œ œ œ œ œ œ œJ œ ™ Œ œ ˙ œ Œ Ó Œ œ œ
˙ ™ œ œ ˙ œJ œ ™ œ œ œ œ œ œ œ œ Œ Ó
Œ œ œ œ œ œ Œ œ ˙ œ Œ
IV
To night- with words un spo
iii
ken,-
IV
you said that I'm the on ly- one.
I
IV
But will my heart be bro
iii
- ken when the
night
vi
meets the morn
V/V
in'- sun.
IV V
&1:03(orig. C)
!
&
&
&
Œ œ œ œ ˙ Œ œ œ ™ œJ œ Œ
Œ œ œ œ œ œ œ œ œ œ œ ™ œJ œ ™ œj œ Œ Ó
Œ œ œ œ ˙ Œ œ œ ™ œJ œ Œ Ó Œ œ œ
œJ œ ‰ Œ œ œ ˙ ™ œ ˙ œ œ œ ˙ ™ Œ
Chapter 4: Conversions 182
Example 4.4.03: “Will You Love Me Tomorrow” (The Shirelles, 1960); form chart
Start Mm. Lyrics Sub. Part Group 0:00 4 ---- (vamp) intro 0:07 8 “Tonight you’re mine completely....” SR 0:21 8 “Tonight, the light of love is....” DC
A1
0:35 8 “Is this a lasting treasure....” SR 0:49 8 “Can I believe the magic....” DC
A2
1:03 8 “Tonight with words unspoken....” B1 1:17 8 “But will my heart be broken....” B2
B
1:31 8 “I’d like to know that your love....” SR 1:45 8 “So tell me now and I won’t ask....” DC
A3
AABA
1:59 8 ---- (solo on SR harmonies) (SR) solo 2:13 8 “So tell me now and I won’t ask....” DC A3’ 2:27 ~ 8 “Will you still love me....” ref. outro
Covach uses this song as an exemplar of AABA form (2005, 71). Indeed, this song
presents many of the hallmarks of AABA forms in which a 16-bar SRDC structure comprises
the A material. As one factor, the 16-bar SRDC structure itself is very typical: the S and R
gestures begin in identical ways, with the restatement ending on the dominant; the departure
gesture includes shorter melodic phrase lengths as well as harmonies that move away from
the major-mode tonic; and the conclusion gesture includes a prototypical tail refrain. The B
material of the song also seems like a classic bridge section, most obviously because the
harmonies reflect an underlying S–T–S–D pattern. From the perspective of lyrics, there is
not much repetition of text within the 64-bar core AABA structure aside from the tail refrain.
As a result, we might judge the A sections to be quite verse-like throughout. That being said,
there does seem to be a schism between the SR and DC gestures (as reflected in Example
4.4.03), in part because of the entrance of the background vocals in the initial DC gesture as
well as because of the way the SR and DC gestures are treated differently (one with solo, one
with vocals) near the end of the song. Temperley, in fact, labels these SR and DC gestures as
verse and chorus sections, respectively (2010). Overall, we can see that this song raises the
same kinds of issues with regard to section labels that we explored earlier. Consequently,
these longer AABA forms will not be the focus of the following discussion.
Historic AABA forms: A section as chorus, B as bridge
A different set of issues arises when we investigate AABA patterns in which the A
material is shorter than 16 bars. In fact, AABA forms as long as 64 bars may not be the
clearest examples of AABA form, since their A sections potentially split into multiple section
roles. Indeed, when Covach presents AABA form, he derives it from an earlier (pre-rock) 32-
bar scheme (2005, 69). It thus appears that the 32-bar framework may be a clearer model for
a “non-compound” AABA form.
Chapter 4: Conversions 183
The 32-bar AABA form was a common structure in American popular song during the
first half of the 20th century, i.e., the Tin Pan Alley years (Covach 2005, 69). Yet the
relationship between this older 32-bar AABA structure and AABA form in rock music is not
entirely straightforward, as section label usage differs between the two eras. For instance,
both Covach and Stephenson present the song “Over the Rainbow” (by E. Y. Harburg and
Harold Arlen, 1938) as a classic example of the old-style 32-bar AABA format. Interestingly,
musicians from the Tin Pan Alley era referred to this entire 32-bar AABA span as the “chorus”
of the song. Yet no theorist of rock music has presented a labeling scheme in which the entire
AABA form is referred to as a chorus. So while the older 32-bar AABA form seems to have
musically influenced form types in rock music, the labeling scheme has not necessarily
followed. To differentiate between this earlier usage of the term “chorus” and its more
modern meaning, Covach refers to the 32-bar AABA form as a “sectional chorus” (2009, 26).
Doing so, Covach can adopt his standard practice of labeling the A section of the AABA form
as a verse and the B section as a bridge without the apparent conflict of having a verse inside a
chorus.
It should be noted that the original version of “Over the Rainbow” also includes an
additional section that precedes the famous 32-bar AABA sectional chorus. This part
(beginning with the words “When all the world is a hopeless jumble”) was referred to at the
time as the “verse” of the song (Stephenson 2002, 136). (N.B., This verse is not part of Judy
Garland’s famous recording of the song.) Many songs from the Tin Pan Alley era, in fact,
include both a “sectional verse” and a “sectional chorus” (to use Covach’s terminology).
Typically, the sectional verse is a forgotten appendage of the song, and it is the 32-bar AABA-
structured sectional chorus that we remember. In this regard, earlier usages of the terms
“verse” and “chorus” appear to mirror modern usages, as the chorus (or sectional chorus) is
clearly the more memorable part of the song. Consequently, we may not be able to entirely
discount the overlap in terminology between the two eras.
The chorus-like quality of the entire 32-bar AABA pattern in songs from the Tin Pan
Alley era is, in fact, a central part of why Stephenson uses the chorus label for the A sections
of AABA patterns in rock songs. Of AABA patterns in rock songs, he writes: “since the
[sectional] verse is absent and the chorus, in the old sense of the word, constitutes the whole
song, it may be better to use the word chorus to refer only to the most repeated musical
passage [i.e., the A section]….” (2002, 137). This practice may seem somewhat confusing, as
the term “chorus” has shifted from encompassing the entire 32-bar structure to just the A part
of this structure. Nevertheless, we find other evidence that chorus quality adheres to the A
sections of AABA patterns within the descriptions of other authors as well. For instance,
Covach states that in a verse-chorus form, the focus is on the chorus, while in an AABA form,
Chapter 4: Conversions 184
the focus is on the verse (i.e., the A section) (2005, 71). If we associate focal quality with the
term “chorus,” it is not unreasonable to then assume that some similarity exists between the
chorus of a verse-chorus song and the A section of an AABA form. In this regard, the term
“verse” for the A section of an AABA song may not always be well-suited.
In fact, a number of rock songs in AABA form seem to challenge the notion that the A
material functions as a verse. For instance, consider the song “Love Me Do” (The Beatles,
1963). As shown in Example 4.4.04, the song has a clear AABA pattern of parts. As we listen
to the song, the A section seems to have a strong chorus-like quality. As one reason, each
iteration of the A section begins with the title lyrics. Moreover, this A material (Example
4.4.05) consists mostly of short melodic phrase fragments, similar to those we have seen
associated with other chorus sections (e.g., “Shake, Rattle and Roll”). The rhyme scheme
takes advantage of the close spacing of these phrase fragments, drawing our attention to this
section. As well, the title lyric reappears as part of an extended tail refrain at the end of the
section. But perhaps the strongest evidence for chorus quality in this song is that the musical
and lyrical content of each A section is identical. Consequently, these A sections are
undeniably the most memorable and most focal moments of the song. A verse label thus
seems wholly mismatched to our perception of the role played by the A sections in this song.
Rather, the song form appears to be a sequence of choruses interrupted by a lone bridge.
(The song “Mama Said” [The Shirelles, 1961] can be viewed as a similar case.) The chorus-
like quality of A section material will be an important factor in other examples discussed
below.
Example 4.4.04: “Love Me Do” (The Beatles, 1963); form chart
Start Mm. Lyrics Part Group 0:00 8 ---- intro 0:13 13 “Love, love me do....” A 0:34 13 “Love, love me do....” A 0:56 8 “Someone to love, somebody new....” B 1:09 13 “Love, love me do....” A
AABA
1:30 12 ---- solo 1:49 13 “Love, love me do....” A 2:10 4+ ---- outro
Chapter 4: Conversions 185
Example 4.4.05: “Love Me Do” (The Beatles, 1963); A section
Classic 32-bar AABA form
Overall, we can say that some sort of connection exists between the use of AABA
forms in the first half of the 20th century and the rock era. In fact, the 32-bar framework of
early AABA forms proves to be a useful reference point from which to view conversions of
AABA form. On the simplest level, one can easily imagine replacing the 8-bar A section with
longer spans of music, such as a 12-bar blues or a 16-bar SRDC structure, to create larger and
larger structures. This technique is the basic process described in Covach 2005 (70).
Deriving compound AABA forms (with clear verse and chorus sections) from the simple 32-
bar AABA form becomes a straightforward procedure of filling the A sections with longer
spans of music.
Other changes to the 32-bar AABA form create a different set of interactions with
section roles. To more fully appreciate these situations, it is helpful to have a more detailed
reference point for AABA form than simply the letter sequence itself. (As the reader should
recall, specific harmonic and melodic organizations helped track conversions of 12-bar blues
and 16-bar SRDC patterns.) In this regard, the song “I’m Walkin’” (Fats Domino, 1957) is
worth examining, since it displays a very common structure for 32-bar AABA forms prevalent
in the 1950s and 1960s. As Example 4.4.06 shows, the song has a clear 32-bar AABA core,
which occurs multiple times during the song. More importantly, though, the harmonic and
melodic organization of the AABA core is typical of many 32-bar AABA songs in rock. The B
section (Example 4.4.07), for example, presents a classic bridge section. Bridge quality for
this B section derives strongly from the S–T–S–D harmonic background, especially its
Example 4.4.06: “I’m Walkin’” (Fats Domino, 1957); form chart
Start Mm. Lyrics Part Group 0:00 4 ---- intro intro 0:05 8 “I’m walkin’, yes indeed....” A1 0:14 8 “I’m lonely as I can be....” A2 0:22 8 “What you gonna do when the well run dry....” B 0:31 8 “I’m walkin’, yes indeed....” A1
AABA
0:40 32 ---- (over AABA harmonies) solo solo 1:15 8 “I’m walkin’, yes indeed....” A1 1:23 8 “I’m lonely as I can be....” A2 1:32 8 “What you gonna do when the well run dry....” B 1:40 8 “I’m walkin’, yes indeed....” A1
AABA
1:49 16 ---- (fadeout on AA....) solo outro
Example 4.4.07: “I’m Walkin’” (Fats Domino, 1957); B section
Example 4.4.08: “I’m Walkin’” (Fats Domino, 1957); A section
The A section of the song (Example 4.4.08) also presents a typical construction,
which will be referred to here as a classic 8-bar A section. This scheme is characterized by
numerous features. In particular, the first two bars contain a melodic motive, which is
followed in the next two bars by a restatement or response to this opening motive. The
connection between these two melodic motives is reinforced through a rhyme in the lyrics
found here – in which the I–iii–IV–V progression is repeated – affirms this underlying
statement-restatement organization. The prototypical tail refrain that ends this second A
section further confirms its classic 8-bar A section structure.
Example 4.4.11: “True Love Ways” (Buddy Holly, 1960); organization of consequent A section
The first A section, however, departs somewhat from a classic 8-bar A section
scheme. For example, the melodic fragments in the second 4-bar hypermeasure are not the
long concluding phrase of a prototypical tail refrain. That being said, the break between the
two short vocal phrases (“Will by and by” / “know true love ways”) is relatively small
compared to other vocal breaks in the song, and thus it is not hard to imagine an overarching
melodic grouping (as shown via the dotted phrase marking in Example 4.4.10). The more
obvious difference is in the seventh bar, where the typical cadence on tonic is eschewed.
Instead, the harmonic progression comes to rest in the eighth bar on a dominant chord,
creating the feeling of a half cadence. A half cadence at this point in an AABA form is a
feature that has been noted by other theorists (e.g., Stephan-Robinson 2009, 161-162). But
this half cadence is not simply an attribute of classic 32-bar AABA form. Rather, it is a critical
alteration in the conversion of classic AABA structures. Instead of two separate, closed
sections (A and A), the addition of the half cadence links these two sections into something
more like a single, large 16-bar unit. One could say, in fact, that these two A sections exhibit
what theorists of common-practice music would call a “period” form (e.g., Caplin 1998, 49),
in which the first A section functions as an antecedent and the second A section functions as a
consequent. For now, though, a more in-depth discussion of how this change potentially
affects our sense of section roles will be tabled.
A few final insights are worth noting with regard to “True Love Ways.” The B section
of this song (shown in the first half of Example 4.4.12) generally conveys a strong bridge-like
quality. In a typical fashion, the tonic chord is in a relatively weak hypermetric position; as
well, the section begins on subdominant and moves to a dominant chord, which is intensified
via its own dominant.
I V I& 1̂ !ŒÓ Œ ŒÓ Œ ŒÓ
Chapter 4: Conversions 190
Example 4.4.12: “True Love Ways” (Buddy Holly, 1960); B and A sections
Example 4.4.13: “True Love Ways” (Buddy Holly, 1960); form chart
Start Mm. Lyrics Part Group 0:00 8 “Just you know why....” A1 0:25 8 “Sometimes we’ll sigh....” A2 0:48 8 “Throughout the days....” B 1:11 8 “Sometimes we’ll sigh....” A2
AABA
1:34 8 ---- (over A2 harmonies) solo solo 1:57 8 “Throughout the days....” B 2:20 8+ “Sometimes we’ll sigh....” A2
BA
Bridge-like quality is also conveyed through the succession pattern of the song (see
Example 4.4.13). The song has a clear AABA core, followed by an instrumental solo, and then
a repeat of the BA material. This distribution of parts (AABA, instrumental break, BA) is a
prototypical large-scale succession pattern for classic 32-bar AABA songs – so much so that
all five of the classic AABA examples from 1957 mentioned above have this overall form. The
placement of the B section in the large-scale succession pattern of the piece thus helps convey
its bridge-like quality. With regard to the quality of the A section, it is once again not entirely
clear whether a verse or chorus label might be more appropriate. As seen in “I’m Walkin’,”
the lyric structure provides evidence to support both readings, and – lacking any other
candidates – little else in the song helps in the decision process. Finally, note that the B
this musical material results from an extension of the classic 12-bar blues can be found just
before the instrumental solo (around 1:07), where an extra bar is added to this 12-bar unit to
give us the standard two bars of rest before the beginning of the next section. Overall,
evidence for a classic 8-bar A section appears to be absent here. Yet a classic 8-bar A section
and a classic 12-bar blues are not so different. Both structures present a melodic motive
grounded strongly in tonic harmony, followed by a repeat or response to this melodic motive,
followed by a prototypical tail refrain. The main difference is that, in a classic 8-bar A
section, the motives are contained within two 2-bar spans, whereas within a classic 12-bar
blues, the motives are contained within two 4-bar spans. The close similarity between these
two form types may help explain why the classic 12-bar blues so often acts as the A material in
an AABA form.
If we look at the large-scale form of the song (Example 4.4.15), we see that this 12-bar
blues unit indeed acts as the A material within a typical AABA succession strategy. Ignoring
for now the opening and closing sections (marked “B-alt”), we find here as well the standard
AABA–break–BA formula so common to classic 32-bar AABA songs. In fact, the original
arrangement of this song did not have an opening B section. Instead, the song launched
immediately into the AABA–break–BA pattern. Only in production did George Martin (the
famed producer of The Beatles) suggest that the song begin with an altered version of the B
section (Martin 1979, 133). The classic AABA roots of this song thus seem relatively clear,
both in terms of large-scale form as well as the harmonic and melodic organization of the A
sections.
Example 4.4.15: “Can’t Buy Me Love” (The Beatles, 1964); form chart
Start Mm. Lyrics Part Group Alt. 0:00 6 “Can’t buy me love....” B-alt intro chorus 0:09 12 “I’ll buy you a diamond ring....” A1 verse 0:26 12 “I’ll give you all I’ve got....” A2 verse 0:42 8 “Can’t buy me love....” B chorus 0:53 12+1 “Say you don’t need no diamond....” A3
AABA
verse 1:12 12 ---- (over A section harmonies) solo 1:29 8 “Can’t buy me love....” B chorus 1:40 12 “Say you don’t need no diamond....” A3
BA verse
1:57 8 “Can’t buy me love....” B-alt outro chorus
Chapter 4: Conversions 193
Example 4.4.16: “Can’t Buy Me Love” (The Beatles, 1964); B section
The B section of this song does not clearly act as a bridge, however. In fact, George
Martin refers to this B section as the chorus (1979, 133). As we examine this B section
(Example 4.4.16), we can appreciate its ambiguous role within the song. The harmonic
structure, for example, conveys a strong sense of classic bridge quality. As one factor, the B
section moves strongly to a retransitional dominant chord at its end. Also, its 8-bar length
accords with our notion of classic bridge sections. While the opening iii–vi progression may
not immediately jibe with our notion of bridge openings, the emphasis on the submediant is a
common enough substitution for a subdominant harmony (as seen in SRDC patterns). Yet
certain important features exist within this B section that strengthen its chorus-like quality.
The use of a submediant chord instead of a subdominant (as noted in the discussion of DC
gestures) may be seen to give the section a somewhat more stable harmonic quality overall.
More importantly, the vocal melody includes what appear to be two head refrains at the
beginning of each 4-bar hypermeasure. These end-accented, title-containing, title-repeating
phrases strongly demand our attention. Part of the effectiveness of these head refrains is the
way in which they contrast with the other melodic material in the song. In this regard, note
how the melody of the A section (or verse) is fairly flat-footed, metrically speaking. The verse
melody begins rather squarely on the downbeat and continues using strong hyperbeats as a
starting point. With the beginning of the B section, though, the melodic phrase rhythm
dramatically shifts forward in time (from the beginning-accented melodic phrases in the verse
section to the end-accented melodic phrases in the chorus). This forward shift calls attention
to itself, as it disturbs the more regular interaction between the melodic grouping and meter
as seen in the A section. In the third measure of the B section, we hear a shift back to the flat-
footed melodic organization of the verse. This shift allows the next melodic phrase (directed
towards the downbeat of the fifth bar) to achieve the same perceptual salience as it shifts back
again to the end-accented strategy. George Martin must have been keenly aware of the
chorus-like quality of the B section to make the suggestion that this part open the song as a
the material at 1:45 sounds a lot like the bridge section of the song as a whole (even though it
arguably ends on a strong tonic chord – atypical for a bridge – before the final dominant
turnaround). The bridge-like quality of this new material heightens our sense that the B
material – which may have seemed somewhat bridge-like – acts not in a bridge role but
rather as a chorus.
Example 4.4.22: “Suspicious Minds” (Elvis Presley, 1969); form chart
Start Mm. Lyrics Part Group Alt. 0:00 2 --- intro 0:04 8 “We’re caught in a trap....” A1 0:21 8 “Why can’t you see....” A2 0:38 8 “We can’t go on together....” B
Vr-Ch
0:55 8 “So with an old friend I know....” A3
AABA
1:11 8 “But here we go again....” A4 1:28 8 “We can’t go on together....” B
Vr-Ch
1:45 12 “Oh, let our love survive....” Bridge 2:15 8 “We’re caught in a trap....” A1 2:31 8 “Why can’t you see....” A2
Vr AA
2:48 8 (6x) “.... caught in a trap....” (fade) outro
The conversion of a classic 32-bar AABA form into a clear verse-chorus form may be
so complete here that few remnants of AABA form may be apparent to a listener. Yet the
AABA roots of this verse-chorus song can help explain interesting features of its construction.
In particular, note the unique way in which this song ends. At 2:15, the verse-like material
returns after the modern bridge. This post-bridge return to verse material is certainly not a
surprising turn of events in the context of a verse-chorus song. What is surprising is that the
chorus never returns after the modern bridge. Instead, the song presents numerous
iterations of the opening musical material. These iterations eventually fade out, but not
before we hear multiple repeats performed at high levels of dynamic intensity. The song does
not end, in fact, until around the 4:30 mark. This means that basically half of the song (from
2:15 to 4:30) consists of repetitions of the opening 8-bar unit, most of which include identical
lyric content. By the third or fourth iteration of this 8-bar unit, the listener may very well
begin to feel that the repeated section (“We’re caught in a trap”) is actually the most
important or most focal part of the entire song. The apparent focal quality of this music
should be somewhat understandable, though. As discussed above, focal quality – if not
chorus quality – is something that theorists associate strongly with the A sections of classic
32-bar AABA organizational schemes. Thus while the A and B sections in the first half of this
song act in a relatively clear verse-chorus relationship, some focal quality still adheres to the
Chapter 4: Conversions 200
A material. It is this focal quality – a weak trace of the classic 32-bar AABA heritage from
which the form of this song seems to derive – that is revealed at the end of the song.
Verses versus verses
“Suspicious Minds” raises another important issue with regard to song form – in
particular, the issue of what exactly constitutes “the verse” of the song. With a verse-chorus
reading of “Suspicious Minds,” one would traditionally consider the entire 16-bar span from
0:04 to the beginning of the chorus at 0:38 as “the verse.” As we saw, however, there were
underlying elements of classic 32-bar AABA structure, and it was not difficult to hear the first
32 bars of the song as a standalone AABA form. With this reading, each 8-bar A section
would traditionally be considered “the verse” material of the song. (See, for example, the
AABA analyses in Covach 2005.) It may seem somewhat trivial to debate whether the first 16
bars of vocal material should be considered as one verse or two successive verses. Indeed,
this distinction is not necessarily valuable in and of itself. Nonetheless, we should recognize
that what might seem like a single verse in one analysis may be two individual verses in
another. In the past, theorists have made statements to the effect that the AABA and verse-
chorus forms are incompatible, e.g., “AABA form is fundamentally different from…. verse-
chorus form” (Stephan-Robinson 2009, 125) or “The strategy of a verse-chorus song differs in
a fundamental way from that of an AABA tune” (Covach 2005, 71). Part of this
incompatibility may derive from the difference in how verse lengths are conceived. “The
verse” of an AABA form does not necessarily equal “the verse” in a verse-chorus form. Being
able to recognize this difference helps us navigate some of the more subtle relationships in
song forms.
In short, different conceptions of verse lengths can obscure relationships between
AABA and verse-chorus forms. This situation becomes most extreme when the verse and
chorus lengths are the same size and exist within a standard verse-chorus succession pattern.
The song “Sin City” (The Flying Burrito Brothers, 1969) provides a useful illustration of this
issue. Listening to the song, it should be obvious that we have a 16-bar verse (Example
4.4.23) followed by a 16-bar chorus (Example 4.4.24). This verse-chorus structure is
confirmed through the pattern of text repetition, as each verse contains new lyrics and each
chorus contains the same lyrics. A verse-chorus reading is also confirmed by the succession
of sections in the song, which comprises a core pattern of Vr–Ch–Vr–Ch–solo–Vr–Ch. As
well, the fact that the chorus and verse sections are the same size makes the 16-bar length
seem like the fundamental unit of the song. For these reasons, this song appears to be an
exemplar of verse-chorus form.
Chapter 4: Conversions 201
Example 4.4.23: “Sin City” (The Flying Burrito Brothers, 1969); verse material
Example 4.4.24: “Sin City” (The Flying Burrito Brothers, 1969); chorus material
Upon closer examination, however, the song displays many attributes of a classic 32-
bar AABA organizational scheme. For example, the verse section can be considered – via its
Example 4.4.26: “Handy Man” (Jimmy Jones, 1960); B and A sections
Example 4.4.27: “Handy Man” (Jimmy Jones, 1960); form chart
Start Mm. Lyrics Pt. Group 0:00 4 “Come-a, come-a, come-a....” intro 0:07 8 “Hey girls, gather round....” A1 0:20 8 “I’m not the kind to use....” A2 0:34 8 “If your broken hearts....” B 0:46 8 “Here is the main thing....” A3
AABA
1:00 8 --- solo 1:14 8 “If your broken hearts....” B 1:25 8 “Here is the main thing....” A3
BA
1:41 8+ “Come-a, come-a, come-a....” outro
Although this Jimmy Jones version of “Handy Man” is a relatively clear example of a
classic 32-bar AABA organizational scheme, one interesting wrinkle is important to mention.
In particular, note how the melody in the post-bridge A section (A3) does not begin as do the
melodies of the other A sections (refer to the second half of Example 4.4.26). Instead of
starting on a ^5, the voice descends from a high ^1. The necessity of this alteration derives
from the ending of the B section. As Jones sings “they’ll come runnin’ for me,” he leaps up
into an extremely high register, and the beginning of the A3 section necessarily transitions
between this high ending and the more moderate vocal tessitura of the A section. The
If your
IV
bro ken- hearts need re pair,- I
I
'm- the man to see. I
whis
IV
per- sweet things you tell all your friends; they'll
In Taylor’s version, however, a critical change occurs in the last A section of the AABA
pattern. Like Jones, Taylor descends from a high tonic pitch at the beginning of this A3
section. But Taylor underpins this melodic descent with IV–V–I harmonies instead of just
tonic harmony. This seemingly small change potentially creates a dramatic effect on our
perception of section roles in this song. Instead of sounding like a return to the opening
material, this altered A section sounds like something different. The A3 section feels more
like an ending – more like it drives to the eventual cadence on tonic in its seventh bar. There
is something distinctly chorus-like about the A3 section as a result, in that this section more
clearly acts as a closing part to the larger whole. At minimum, the “pre-B” and “post-B” A
sections are not obviously the same, and many casual listeners may not realize that any return
to the A material whatsoever occurs immediately after the B material.
The standalone and chorus-like quality of the A3 section is emphasized in the
succession pattern of the song as well. Unlike in Jones’s version, the succession pattern in
Taylor’s version (Example 4.4.30) does not include a separate instrumental break and B
section prior to the last A3 section. Instead, the 8-bar unit starting at 1:46 merges the first six
bars of intro material from Jones’s versions (“Come-a, come-a, come-a”) with the last two
bars of the B section (“They’ll come runnin’ to me”). With this format, the A3 section seems
even less connected to the AABA pattern as a whole. Since the A3 section at 2:08 basically
repeats the lyrics from its early instance, it also begins to sound somewhat chorus-like.
Interestingly, the one lyric change Taylor makes in this second A3 iteration is that the tail
refrain now includes the title text. Consequently, the focal quality of this final A section
becomes even stronger still. (Note that the Jimmy Jones version does not include the title
text in the second BA iteration.)
Example 4.4.30: “Handy Man” (James Taylor, 1977); form chart
Start Mm. Lyrics Pt. Group 0:00 8 --- intro 0:21 8 “Hey girls, gather round....” A1 0:43 8 “I’m not the kind to use....” A2 1:04 8 “If your broken hearts....” B 1:25 8 “Here is the main thing....” A3
AABA
1:46 8 “Come-a, come-a, come-a....” (intro/B hybrid) 2:08 8 “Here is the main thing....” A3 A 2:29 8 “Come-a, come-a, come-a....” 2:50 8 “Come-a, come-a, come-a....”
outro
Overall, a subtle yet tangible sense exists that the B and A3 sections in Taylor’s
version act as prechorus and chorus material, respectively speaking. Admittedly, these
section roles are not entirely clear. In fact, it would be possible to make an argument that the
Chapter 4: Conversions 207
“Come-a, come-a, come-a” sections have a certain chorus-like quality about them as well.
Nonetheless, the differences between Jones’s and Taylor’s versions show an important step in
the continuum between AABA and verse-chorus forms. In particular, the unstable nature of
typical B-section material focuses our attention on the material that follows. In this regard, a
“post-B” A section has the potential to become more chorus-like than earlier iterations of A
sections in a song. As the “post-B” A-section material becomes more chorus-like, our
perception of the B material as something closer to a prechorus becomes stronger as well.
One final example – the song “God Save the Queen” (The Sex Pistols, 1977) –
provides an even clearer illustration of this effect. This song is a particularly good example
because the “post-B” A section seems to be conceived as a chorus by the songwriters
themselves. In Example 4.4.31, the large-scale succession pattern for the song is mapped out.
As can be seen, the song clearly divides into separate 8-bar units. More importantly, there
appears to be some evidence of 32-bar AABA thinking, as these 8-bar units consistently
organize into an AABA pattern (as the “Group” column shows).
Example 4.4.31: “God Save the Queen” (The Sex Pistols, 1977); form chart
Start Mm. Lyrics Part Group 0:00 4 ---- vamp 0:06 8 ---- (using A1 harmonies) link
intro
0:19 8 “GSTQ, the fascist regime....” A1 0:32 8 “GSTQ, she ain’t no human being....” A2 0:44 8 “Don’t be told what you want....” B1 0:57 8 “GSTQ, we mean it man....” A3
AABA
1:10 8 “GSTQ, ‘cause tourists are money....” A4 1:23 8 “God save history....” A5 1:36 8 “When there’s no future....” B2 1:49 8 “GSTQ, we mean it man....” A3
AABA
2:02 8 ---- (over A1-like harmonies in F#) solo solo 2:17 8 “GSTQ, we mean it man....” (alt.) A3’ A3’ 2:30 8 “No future, no future....” C 2:43 8 “No future, no future....” C 2:56 8+ “No future, no future....” C
outro
In a general way, the internal characteristics of the 8-bar units corroborate this 32-
bar AABA structure. For example, the opening vocal material (Example 4.4.32) displays
similar qualities to a classic 8-bar A section. Tonic harmony is basically prolonged here, and
we also find two parallel, rhyming vocal phrase fragments in the first 4-bar hypermeasure.
Although a prototypical tail refrain is absent, there does seem to be some sort of a coming-to-
rest in the seventh bar, as the harmonic motion pauses on a tonic chord. The B material (not
shown) is somewhat different than a classic bridge, in that only two chords – E and B major –
comprise its harmonic content. In the context of the prevailing tonal center of A major, we
Chapter 4: Conversions 208
would label these E and B major chords as V and V/V respectively. (In isolation, though, the
B section sounds as if it is in E major, so Roman numerals I and V might be more
appropriate.) In this regard, the B material can be seen to basically prolong an E major
harmony and thus act as an unstable harmonic area in the tonal context of the A sections.
Classic A section construction is more evident within the A3 part (Example 4.4.33). Now,
prototypical tail refrain quality becomes more obvious, as the addition of a dominant chord
creates a strong drive to the cadence in the seventh bar. As well, the vocal melody comes to a
rest closer to the seventh bar downbeat than in the A1 or A2 sections. Overall, the internal
characteristics of these 8-bar units align with our general expectations of typical A and B
sections in a 32-bar AABA organizational scheme.
Example 4.4.32: “God Save the Queen” (The Sex Pistols, 1977); A1 section
Example 4.4.33: “God Save the Queen” (The Sex Pistols, 1977); A3 section
From a number of perspectives, though, it appears as if the A3 section should be
considered to fulfill a separate role than the A1 and A2 sections. Obviously, the differences in
harmonic and melodic organization between the “pre-B” and “post-B” sections encourage us
to hear them in different section roles. Further differences can be found, though. Note how –
God
I
save the queen,
IV
the fas
I
cist- re gime.-
IV
They
made
I
you a mor on,-
IV
po ten
I
- tial- H bomb.-
&0:19(orig. A)
~~~~ ~~~~~~
& ~~~~ ~~~~~ !
œ œ œb œb œ Ó Œ ‰ œj œ œ œ œb ™ œj Ó Œ ‰ œb j
œn œ œ œ œ# œ# œj ‰ Œ ‰ œb j œb œn œ œ œ
God
I
save the queen,
IV
we mean
I
it, man.
IV
We
love
I
our queen.
V
God
I
saves.
&0:57
(orig. A)
~~~~
& ~~~~~ ~~~~~~~~~~
œ# œn œb œ# œ Ó Œ ‰ œn j œj œb ™ ˙ Ó Œ ‰ œb j
œj œ ™ œ œ Ó Œ œb œj œn ™ ˙ œj‰ Œ Ó
Chapter 4: Conversions 209
despite that fact that every A-like section begins with the title phrase – the lyrics to the A3
section are the same on its first and second iterations. In contrast, the post-title lyrics in the
other A-like sections are constantly changing. In this regard, the A3 section seems to be
conceived as the chorus section of the song. Further evidence of this chorus-like aspect can
be found in the domain of instrumentation. The “pre-B” sections are notably restrained in
terms of dynamics: we hear palm-muted chords in the electric guitar, and the drummer plays
mostly a closed hi-hat. In contrast, the “post-B” sections are much thicker sounding: the
electric guitar chords are now fully strummed, and the drummer moves over to the ride
cymbal for a bigger sound. The combination of these factors draws our attention to the A3
section as the focal moment in the song. As a final factor, consider the large-scale grouping of
sections. Having two complete AABA iterations, followed by a solo and a return of the A
material, is a rare format for an AABA song. Rather, we see evidence of a “compound” AABA
form (here, AABA’), which we associate with a large-scale grouping structure for verse-chorus
material.
Overall, “God Save the Queen” seems to have some A-like sections that act as verse
material and others that act as chorus material. In this regard, the B material – which might
otherwise be considered a bridge – seems to take on the role of prechorus as it prepares for
the arrival of the chorus-like A3 sections. We thus have a form that – while highly similar to a
classic 32-bar AABA scheme – exhibits apparent verse-prechorus-chorus construction. The
verse, prechorus, chorus qualities within “God Save the Queen” are admittedly weak in
comparison to other song examples. It is not too difficult, however, to extend the conversion
process seen in this example to other songs. Many songs with a clear prechorus section, for
instance, use the same harmonic content within the chorus section as the verse. Good
examples of this situation include “End of the Road” (Boyz II Men, 1992), “Feels Like the First
Time” (Foreigner, 1977), “Billie Jean” (Michael Jackson, 1983), and “Cool It Now” (New
Edition, 1984). From a harmonic perspective, these songs have highly similar “pre-B” and
“post-B” material (where the “B” section is the prechorus). But through changes in various
domains other than harmony, the “post-B” material clearly acts as the chorus material, not a
return to the verse.
Conclusion
In the preceding discussion of AABA forms, we primarily investigated one single type
of AABA structure – classic 32-bar AABA form – and how the parts of this form can become
converted to take on different roles. As we have seen, the way in which classic AABA form
interacts with verse-chorus form are numerous. For instance, the A section might be
expanded to include both verse and chorus material; or the B section might – in concert with
Chapter 4: Conversions 210
hypermetric reinterpretation – take on the role of chorus; additionally, the B section may
appear to act as a prechorus in preparation for a final A section that has taken on a chorus-
like role. As in the case of blues and SRDC organizational schemes, we can chart out the
changing relationships between section roles and conversions of the 32-bar AABA
organizational scheme. Example 4.4.34 gives such a chart.
Example 4.4.34: Relationships between AABA-derived schemes and section roles
Unlike similar charts presented earlier in this chapter, each box containing an upper-
case letter now represents an 8-bar unit. The classic 32-bar AABA form is shown in Example
4.4.34 and we can see that this organizational scheme spawns numerous branches. In one
lineage, a hypermetric reinterpretation creates the configuration shown in c), as found in the
song “Blitzkrieg Bop.” From this configuration, we move to d), in which the chorus-like
quality of the B material becomes emphasized (as in “Suspicious Minds”). The path to f), in
Verse Verse VerseBridge
AA B A
a)
Verse Chorus
AA B B
f)
Chorus Chorus ChorusBridge
AA B A
b)
ChorusPrechorusVerse
AA B A
e)
ChorusPrechorusVerse
AA B A A
g)
Verse Verse Bridge
AA B
c)
Verse Chorus
AA B
d)
Chapter 4: Conversions 211
which we have similarly-sized verse and chorus sections (as in “Sin City”), can be seen as the
natural outgrowth of these changes.
An alternative path from the 32-bar AABA organizational scheme is shown on the
right-hand side of the chart. To begin with, we should recognize the chorus-like quality
inherent in the A material of many AABA songs, as shown in b) (e.g., “She Loves You”). If we
consider that the A section thus may evince either verse or chorus quality, we can understand
the configuration shown in e). This configuration (seen in songs such as “God Save the
Queen” or James Taylor’s version of “Handy Man”) also reveals the close relationship
between a classic bridge section and a prechorus role. From this configuration, the standard
verse-prechorus-chorus form, shown in g), can be seen to arise through a simple expansion of
the chorus material.
Of course, theorists view AABA form within a variety of settings aside from the classic
32-bar scheme. Other instantiations of the generic AABA pattern potentially interact with
section labels in additional ways. Overall, AABA form encompasses a large network of
relationships between specific musical structures and the section labels that attempt to
describe the role these structures play within the form of a song. Nevertheless, the
connections between AABA form and verse-chorus form appear to be much closer than has
been previously discussed within the existing theoretical literature.
4.5: Summary
In this chapter, we have seen how section roles interact with a variety of differently-
sized musical units. Section roles were discussed, for example, in the context of 4-bar units
(the individual gestures of a 16-bar SRDC pattern), 8-bar units (the individual parts of a 32-
bar AABA pattern), 12-bar units (the classic blues structure), and 16-bar units (the SRDC
pattern and half of the AABA pattern). We have thus moved from a bottom level of the formal
hierarchy up through some of the larger spans of music for which a single section role can
account. As musical spans extend further and further beyond the 16-bar threshold, it
becomes harder and harder to posit that only a single section role would be appropriate for
the span as a whole. (How many 64-bar chorus sections can you think of?)
The 12-bar blues, 16-bar SRDC, and 32-bar AABA organizational schemes stood as
useful reference points in this discussion. By focusing on a particular scheme, we could view
how it (or parts of it) could be converted from one section role to another. We found that
modern song forms could develop out of these smaller schemes in many ways. This
developmental process was not necessarily meant to imply any particular history of song
form. Rather, this process gave evidence of the continuum of attributes and attribute
Chapter 4: Conversions 212
strengths that contributes to our perception of form in rock songs. Nonetheless, we saw
potential paths from the song forms commonly used in the early years of rock to those used
more often in more modern songs. A necessary limit in this regard was that the song
examples in this chapter were primarily drawn from earlier decades in the history of rock. In
the following chapter, we will shift gears to investigate the interplay of section roles within
songs from more recent times.
213
Chapter 5: Blends
5.1: Introduction
One of the assumptions in the previous chapter was that our labeling system for
section roles is predicated on the mutual exclusivity of these roles. If we choose to refer to a
section of music as a verse, for example, we are inherently implying that this passage is not a
bridge, chorus, or some other section role. As we found when looking at conversions of blues,
SRDC, and AABA organizational schemes, though, there are often cases in which it is not
clear what section role might best apply to a particular span of music. As a particular span
“converts” from one role to another, there is often some middle stage in which it appears as if
we have evidence for more than a single section role. “Can’t Buy Me Love,” for instance,
seems to stand as a transitional form in the conversion process between AABA and verse-
chorus forms. As a result, the title-containing passage in this song (e.g., starting at 0:42)
appears to contain aspects of both bridge and chorus roles.
There are, in fact, many recurring situations in rock music that challenge
categorization via a single section role label. Yet theorists rarely if ever choose to refer to a
particular passage of music with two section role labels. As this chapter will show, however,
the choice between one label and another often represents a false dilemma, in that a robust
understanding of how we perceive many musical passages necessarily requires us to recognize
multiple roles acting at once.
In this chapter, the term “blend” will be used to describe those situations in which
aspects of two (or more) section roles appear to exist within the same span of music. Like the
term “conversion,” the term “blend” has been borrowed (in a metaphorical way) from the field
of linguistics. In linguistics, a blend is a word that is created through the union of two or
more words (Denham and Lobeck 2009, 197). For example, the word “smog” derives from a
marriage of the words “smoke” and “fog.” Sometimes, a blend will involve taking the
beginning of one word and adding it to the end of another (e.g., “simulcast” derives from the
words “simultaneous” and “broadcast”). In other cases, a blend takes advantage of a sound
that is shared by both words (e.g., “motel” results from the overlap of the “ot” sound in
“motor” and “hotel”). More importantly, a blend represents the combination of two different
concepts (or conceptual categories). The word “smog” does not just combine the words
“smoke” and “fog”; it also combines the separate concepts of “smoke” and “fog” into one
single idea. As conceptual combinations, blends sometimes involve categories that are
traditionally considered to be mutually exclusive within a single labeling system. For
Chapter 5: Blends 214
example, the categories of “breakfast,” “lunch,” and “dinner” are considered mutually
exclusive with regard to the type of meal one is eating. (If someone is said to be eating
breakfast, then we consider that person to not be eating lunch or dinner.) Yet the term
“brunch” stands as a blend (and conceptual combination) of breakfast and lunch. The notion
of a blend can thus be seen to hold great metaphorical power for explaining the combination
of section roles in rock music, since theorists have mostly treated these roles as mutually
exclusive labels within the form of a song.
It was discussed earlier (at the beginning of Chapter 4) that section roles can be seen
as parallel to what William Caplin refers to as “formal functions” in music of the classical era.
In this regard, blends are similar to “form-functional fusion.” For Caplin (1998, 45), form-
functional fusion occurs when two different functions are present in a single group. (A
“group” is simply some span of measures.) The second half of a musical sentence, for
example, fuses both the continuation and cadential functions. In a similar way, we find that
certain passages of rock appear to blend multiple section roles within a single span of music.
Nevertheless, we will not assume that functional fusion and role blends are equivalent
because of the great difference between the styles of music for which these terms apply.
One practical issue is how we should refer to blends of section roles. One seemingly
appropriate technique would be to employ a blend of the role labels themselves. For instance,
the term “vorus” could be used to describe a passage that blends verse and chorus qualities.
But this approach results in some extremely awkward-sounding terms. (Consider “vidge” or
“brerse” as blends of the verse and bridge labels.) As a result, blends of section roles will
instead be referred to using a slash to separate the two terms. A blend of link and chorus
roles, for example, will be referred to as a “link/chorus.” In English, the slash is often used to
represent two seemingly equivalent choices, such as “either/or,” so its usage here to label role
blends seems suitable. The potential for different meanings in the order of role labels before
and after the slash is something that will be utilized at certain points in this chapter. A
“verse/bridge,” for example, will be considered to describe a different situation than a
“bridge/verse.” That being said, the order of section labels when referring to a blend will not
be critical in most cases. For the sake of consistency, though, a secondary role label will
always precede a primary role label (verse, chorus, bridge).
There are a number of different ways that blends can occur in rock music. One
common situation occurs when a passage has strong melodic and harmonic characteristics of
one section role yet the positional characteristics of a different role. Other blends can be seen
to result from differences between local and global aspects of a passage. In some cases, the
blend is simply an ambiguity between section roles. In other cases, the type of amalgamation
is too complex to precisely pin down. As we will see, blends always involve a main section
Chapter 5: Blends 215
role and some other role (whether that be another main role or something else). This aspect
derives in strong part from the central status that the verse, chorus, and bridge labels enjoy
within existing systems of large-scale form categorization (as discussed at the end of Chapter
2). The current chapter will thus be organized around the three main section roles
themselves.
5.2: Verse Blends (part 1)
In Chapter 4, we saw that it is sometimes unclear whether a particular span of music
should be categorized as a single verse section or subdivided into multiple section roles. With
“incipient” verse-chorus forms (such as “Jailhouse Rock”), for instance, it may not be clear
whether the last eight bars of a 16-bar span should be considered as a separate chorus section
or simply more verse material. Outside of clear blues, SRDC, or AABA settings, we find
similar situations where the choice between a separate chorus and “more verse” seems just as
difficult. In his analysis of “Take It Easy” (Eagles, 1972), for example, Covach labels the main
24-measure block of the song (0:17-0:58) entirely as verse material (2009, 355); but he also
admits that the middle phrase of this 24-bar block (starting at 0:31) sounds like it might be a
chorus. This type of ambiguity can be found in songs ranging from “Sunshine of Your Love”
(Cream, 1967) to “Nuthin’ But a ‘G’ Thang” (Dr. Dre, 1992). (Compare analyses in Covach
2009 [287, 554] to Temperley 2010.) If we investigate these kinds of situations more closely,
we find the reasons for this ambiguity to be much the same as those found within the
incipient verse-chorus forms described earlier. Consequently, they will not receive much
discussion here.
There are other situations where it is not entirely clear whether a section acts as a
verse or some different role. Two common cases will be discussed in this chapter. The
simpler case involves the decision between prechorus and verse roles, which will be discussed
below. The more complicated case involves a blend between bridge and verse roles. This
latter case intersects with some other types of section ambiguity, and so its discussion will be
reserved until the end of this chapter.
Prechorus or verse
In the discussion of the prechorus role in Chapter 3, it was noted that theorists
generally take a conservative approach in their application of this label. By far the most
common situation is when an area of apparent prechorus material is labeled simply as “more
verse.” The distinction between blocks of verse-chorus material or blocks of verse-prechorus-
chorus material is, in fact, often quite unclear. The song “Smells Like Teen Spirit” (Nirvana,
Chapter 5: Blends 216
1991) provides an excellent illustration of this situation. To begin with, note that there are
clear areas of verse (Example 5.2.01) and chorus (Example 5.2.02) in this song, and theorists
do not disagree with regard to the labels for these spans of music.
Example 5.2.01: “Smells Like Teen Spirit” (Nirvana, 1991); verse material
Example 5.2.02: “Smells Like Teen Spirit” (Nirvana, 1991); chorus material
Sandwiched in between these clear verse and chorus sections is the musical material
shown in Example 5.2.03. In analyses by both Covach (2009, 512) and Temperley (2010),
this material is considered to be simply more verse material. Christopher Doll, however,
refers to this material as the prechorus section of the song (2011).
The reasons for this analytical disagreement are not too difficult to infer.
Presumably, Covach and Temperley are responding to the lack of harmonic change between
the opening eight bars of the verse section and the eight bars that follow. Our association of
harmonic instability with prechorus quality (e.g., Everett 2009, 146) is such that the
continuation of the 2-bar, tonic-initiated chord progression from the beginning of the verse
into the chorus denies any strong sense that there is a significant departure away from the
verse role. Nevertheless, there are distinct prechorus-like qualities to the music transcribed
As Covach’s chart shows, the solo material can be conceived as basically instrumental
versions of the verse and chorus sections. Undeniably, this is the easiest way of thinking
about this solo material. From this perspective, there is no strong sense that we have
departed from the main section roles of the song, and thus we could simply say that this song
does not contain any bridge material. At the same time, our sense of a bridge role is not
completely absent here. In particular, the location of these instrumental sections coincides
with our expectations for prototypical bridge material: after two iterations of the main
musical material (in this case, the verse-chorus blocks), a contrasting section provides relief,
and this contrasting material is followed by a final iteration of the main musical material.
This large-scale strategy is, of course, the ubiquitous AABA pattern. In this case, however, the
AABA pattern manifests itself primarily within the domain of instrumentation.
(Interestingly, the harmonies in the instrumental verse and chorus sections are somewhat
different from those found in the vocal iterations of this material, and thus we might judge
AABA quality to be subtly evident in the domain of harmony as well.) In other words, the
instrumental versions of the verse and chorus sections act together to create a large-scale B
group. One might argue that a “B group” (or “B section”) in a large-scale AABA pattern and a
“bridge” are not necessarily the same thing. But the qualities of contrast and location –
which are so central to our prototypical notions of the “bridge” category – are integral aspects
Chapter 5: Blends 223
of the B section in any AABA pattern. By viewing these instrumental solo sections as blends
of verse, chorus, solo, and bridge roles, we can offer an explanation as to why this material
occurs in the location that it does within the large-scale form of the song. In other words, the
principle of a bridge can be helpful to explain song form even when a prototypical bridge is
absent.
A significant insight gained from the example of “Smoke on the Water” is that bridge
quality may adhere to section roles that might otherwise appear to be straightforward
instances of non-bridge roles. This insight impacts our conception of form within a number
of songs (and extends beyond the more simple notion of solo and bridge blends, as we will
see). As another example, consider the song “Angel” (Aerosmith, 1987). In this song, two
blocks of verse-prechorus-chorus (VPC) material are followed by an instrumental solo over
harmonies from the chorus, this solo section leads to a bridge section with vocals, and then
chorus material returns to close the song as a whole (see Example 5.3.02).
Example 5.3.02: “Angel” (Aerosmith, 1987); form chart
Start Mm. Section Group Alt. Harmonic content [in Db] 0:03 9 intro (chorus) I . . V | IV . . V | 0:28 8 verse I | | IV | | I | vi | iii | IV | 0:50 8 prechorus V vi | IV | 1:13 9 chorus
A I . . V | IV . . V | vi . . V | IV . . I |
1:38 8 verse % 2:00 8 prechorus % 2:23 8 chorus
%
% 2:45 8 solo (chorus)
A
% 3:07 8 bridge B
B V | IV | I | vi | IV | V | | |
3:30 8 chorus % 3:52 8 chorus
A’ A’ %
4:15 8+ outro (chorus) %
One way of grouping the solo material is shown in the “Group” column of this
example. With this conception, the solo simply prolongs the chorus section that precedes it.
Consequently, the second “A” group extends all the way up until the vocal bridge section at
3:07. At the same time, this instrumental solo section provides great textural contrast to the
blocks of verse, prechorus, and chorus material that have preceded it. Within the domain of
texture, therefore, we can posit a different large-scale grouping, as shown in the “Alt.”
column. In this conception, the bridge role is split between the solo section and the vocal
bridge. We can, in fact, relate the reason for this split to the nature of the vocal bridge itself.
Notice, for example, that the vocal bridge shares a very similar instrumentation to the chorus
section of the song. As a result, the textural relief of the solo section helps the vocal bridge
stand more clearly apart from the chorus sections themselves. (The way that the song comes
Chapter 5: Blends 224
to a halt on the dominant chord around 3:28 can be seen to fulfill a similar purpose at the end
of the vocal bridge.) It is important to note that neither conception of the solo material is
meant to supersede the other. We can say, rather, that the solo section in this song appears to
blend chorus and bridge quality.
We can also find the opposite situation in other songs. Consider, for example,
“Buddy Holly” (Weezer, 1994), a succession pattern for which is shown in Example 5.3.03.
Example 5.3.03: “Buddy Holly” (Weezer, 1994); form chart
Start Mm. Section Group Alt. Harmonic content [in Ab] 0:00 8 verse vi | | I | | 0:16 6 prechorus IV | iii vi | IV | iii vi | IV | iv | 0:28 8 chorus
A I | IV V |
0:43 4 link (verse) vi | | I | | 0:51 8 verse % 1:07 6 prechorus % 1:19 8+2 chorus
A
%
% 1:39 8 bridge B I vi | vi I | 1:55 6+2 solo (prechorus)
B %
2:11 8+6 chorus A’
A’ %
One way of conceptualizing the large-scale form for “Buddy Holly” is reflected in the
“Group” column. This grouping structure accords priority to the harmonic contents of the
various parts of the song. In this reading, the return of the prechorus harmonies in support of
an instrumental solo becomes grouped with the chorus section that follows (since it would in
this case belong to the final “A” group). Yet we might also consider that this instrumental
solo provides textural contrast within the song form as a whole; in this light, we can say that
this instrumental solo section evokes bridge-like quality. Moreover, this solo section directly
follows the vocal bridge of the song, and thus it potentially participates in a broader “B” group
overall, as shown in the “Alt.” column. Comparing this “Alt.” grouping structure with the
“Alt.” grouping structure for “Angel” above, we find a close alignment between the basic
organizational strategies. In both songs, we find two VPC blocks, a B group consisting of a
vocal bridge and instrumental solo material, and then closing chorus material. The ordering
of the vocal bridge and instrumental solo sections is different, and so is the harmonic
relationship of the solo section to other sections in the song. But the general principle
remains the same: a large-scale AABA pattern organizes the song as a whole, while the final A
group is only an abbreviated version of earlier A group material. As in “Angel,” we may also
posit reasons for the disbursement of the bridge role across what appear to be two separate
sections. Specifically, note that the vocal bridge does not depart very strongly from tonic
harmony. In fact, this vocal bridge can be seen to be constructed out of a 2-bar riff that
Chapter 5: Blends 225
begins and ends on tonic. The harmonic departure that one expects of a prototypical bridge
section is fulfilled instead by the instrumental prechorus section that follows. In this regard,
the large-scale bridge role is fulfilled by multiple sections within the song.
The ability of an instrumental solo section to act in a bridge role is a useful concept
even when the instrumental solo seems strongly based on a specific section type. For
instance, consider the role of the solo section in “Smells Like Teen Spirit” (Nirvana, 1991).
One straightforward way of conceiving of the song form is sketched in Example 5.3.05 (which
has been adapted from Covach 2009 [512]). (The verse-prechorus blend described in the
previous portion of this chapter is ignored here for the sake of simplicity.)
Example 5.3.04: “Smells Like Teen Spirit” (Nirvana, 1991); form chart in Covach 2009 (512)
Start Mm. Section Group 0:00 16 introduction 0:33 16 verse 1:06 20 + 12 chorus
example and those above is that the texture in the off-tonic material here is thinned instead of
thickened. Our understanding of the off-tonic material in this song needs to encompass
aspects of at least two section roles, therefore, if not more.
Conclusion
This discussion of chorus blends has hopefully shown that there are many different
types of situations in which we might apply a chorus label. More importantly, these situations
often involve ambiguity with another section role (or roles). In the preceding discussion, we
have found situations in which refrain, link, prechorus, and even bridge quality seems to
merge with our perception of a chorus section. The superficial simplicity of a chorus label can
thus be seen to often hide something more nuanced and complicated going on underneath.
In the following (and final) portion of this chapter, we will see that some of these types of
ambiguity can have significant ramifications for our perception of verse roles as well.
5.5: Verse Blends (part 2)
At the beginning of this chapter, the notion was put forth that a verse label may
sometimes mask evidence of non-verse qualities. One particular case was examined –
specifically, that in which attributes of prechorus quality could be found within spans of
music that might otherwise be conceived of as verse material. We thus came to appreciate the
ambiguity between prechorus and verse roles, as well as the potential to blend one with the
other. Ambiguity between prechorus and verse roles is perhaps the most common and most
simple case of one section role blending with the verse role. A central part of what makes this
case so simple is that the verse label is not intersecting with the other main section types of
chorus and bridge.
We sometimes do find, though, that ambiguity exists between the roles of bridge and
verse or chorus and verse. We saw some evidence of this type of ambiguity in our discussion
of bridge blends, where verse material could be seen to act in a bridge role. Although we will
also see instances of verse and bridge blends below, these situations can be seen as different
from earlier blends between verse and bridge. These types of verse and bridge blends lead to
a special type of ambiguity between verse and chorus roles. Overall, the situations described
below are inherently more complicated than the ambiguity between prechorus and verse
roles, as our conception of the large-scale form of a song can drastically change when
ambiguity exists between main section roles.
Chapter 5: Blends 269
Implications of prechorus/chorus blends
Although the discussion of prechorus/chorus blends was basically concluded in the
preceding portion of this chapter, it is worth revisiting this topic for a moment. In the
preceding discussion, the verse material was never questioned. In other words, the role of the
music preceding the prechorus/chorus blend was clear (and essentially taken for granted) in
every example. Yet this is not always the case.
Consider, for instance, the song “Tears in Heaven” (Eric Clapton, 1992). As a point of
reference, the recent analysis by Temperley (2010) will be used, since it reflects a reasonable
(although not unproblematic) conception of the song form. In Example 5.5.01, we can see the
succession pattern of sections that Temperley posits for this song. Although Temperley does
not offer a higher-level grouping structure in his analysis, we can easily imagine a typical
grouping of these sections based on the labels that he provides. We could say, for instance,
that there are two blocks of verse-chorus material (AA), followed by a bridge (B), after which
we hear two more blocks of verse-chorus material (AA). We could thus posit a large-scale
organization for this song as an AABAA pattern (in which the higher-level bridge role
overlaps – from a textural perspective – the following A group via the guitar solo). All in all,
the form of this song appears to be relatively straightforward, except perhaps that an
additional A group is found after the main AABA presentation. But as one might expect by
now, the form of the song is not as straightforward as this conception might imply.
Example 5.5.01: “Tears in Heaven” (Eric Clapton, 1992); form chart in Temperley 2010
Start Mm. Section Lyrics 0:00 4 intro (link) ---- 0:13 8 verse “Would you know my name....” 0:38 6 chorus “I must be strong and carry on....” 0:56 4 link ---- 1:09 8 verse “Would you hold my hand....” 1:34 6 chorus “I’ll find my way through night and day....” 1:53 4 link ---- 2:05 8 bridge “Time can bring you down....” 2:30 8 solo (verse) ---- 2:55 6 chorus “Beyond the door there’s peace....” 3:14 4 link ---- 3:26 8 verse “Would you know my name....” 3:51 6 chorus “I must be strong and carry on....” 4:10 4 outro (link) ----
To begin with, let us examine the verse and chorus sections that Temperley identifies.
(As in previous examples, the section that Temperley labels as a bridge will be taken for
Chapter 5: Blends 270
granted here in order to simplify the discussion.) Examples 5.5.02 and 5.5.03 show the first
iterations of these verse and chorus sections, respectively. Right away, the reader should
notice the obvious similarity in organization between these two sections and those found in
the examples of prechorus/chorus blends discussed previously. Specifically, we see a span of
tonic-prolonging material, followed by a section of off-tonic material, which – via a refrain –
overlaps into the subsequent link material. The off-tonic quality of Temperley’s chorus
section is especially piquant, as the applied chords – including a supertonic of the supertonic
– heighten the chromaticism of this passage. We might judge the ambiguity in this song to be
explainable simply by the notion of a prechorus and chorus blend. Case closed then, right?
Example 5.5.02: “Tears in Heaven” (Eric Clapton, 1992); Temperley’s verse
Example 5.5.03: “Tears in Heaven” (Eric Clapton, 1992); Temperley’s chorus
Unfortunately, there is a bit of a snag for this reading (verse followed by
prechorus/chorus). In particular, pay attention to the repetition pattern of the lyrics (some of
which are included for reference purposes in the right-hand column of Example 5.5.01).
Surprisingly, perhaps, three of the four iterations of “chorus” material present – aside from
the final refrain – entirely new lyrics. This aspect unarguably compromises our ability to hear
I
Would you know
V6
my name
vi I6$ IV6
if I saw
I6$
you in hea ven?
V
-
I
Would it be
V6
the same
vi I6$ IV6
if I saw
I6$
you in hea ven?
V
-
&(orig. A)0:13
&
Œ œ œ œ œ œ œ œ œ Ó Œ œ œ œ œ œ œ œ Œ Ó
Œ œ œ œ œ œ œ œ œ Ó Œ œ œ œ œ œ œ œ Œ Ó
vi
I must be strong
V6/vi ii 6/ii
and car ry- on
V6/ii
'cause I know
ii
I don't be long-
V1 1
here in hea ven.
I
-
&(orig. A)0:38 o
&
Œ ‰ œJ œ œ œ œ Œ Ó Œ ‰ œbJ œ œ œ œ Œ œ œ œ
œ ‰ œJ œ œ œ œ Œ ‰ œJ œ œ œ
Chapter 5: Blends 271
this section as a chorus, as lyric repetition patterns are a central feature of chorus quality.
That being said, the lack of external lyric repetition is not entirely an issue for our analysis, as
we could say that the more prechorus-like aspects of this prechorus/chorus blend are being
brought to the fore. In fact, this is the exact argument that was used previously in the case of
“Born to Run.”
In “Born to Run,” however, the lyrics in what were considered the verse sections did
not repeat on future iterations either, nor did the verse material include any internal
repetition of lyrics. But that is not the case with “Tears in Heaven.” Temperley’s verse section,
in fact, includes a relatively high level of internal text repetition as well as a significant
amount of text repetition on future appearances. Note how every line in every iteration of this
opening section begins with the word “Would” and ends with the phrase “if I saw you in
heaven?” More than half of the lyrics in these purported verse sections are the exact same
words. There is thus, arguably, some evidence that Temperley’s verse section displays some
chorus-like features.
The chorus-like quality of the first section would not be much of an issue, perhaps, if
this section were followed by a section with strong chorus quality. The hypothetical second
section would so clearly act as the focal point of the song that any focal quality of the first
section would basically be moot. But as we have seen, the second section of “Tears in
Heaven” has a somewhat compromised chorus quality. Aside from the closing refrain, the
lyrics show very low levels of both internal and external repetition. A primary factor for a
verse-chorus reading of these two sections, one might say, is their order within the song. The
link material clearly demarcates separate blocks of material, and – given that two sections
appear to exist within this block – the standard way to parse two such sections is to label the
first as a verse and the second as a chorus. The refrain at the end of the second section is thus
a crucial aspect of the form of this song, for it participates in a coordinated melodic-harmonic
cadence that creates a strong feeling of closure to the larger block as a whole. This large-scale
harmonic closure creates the sense that Temperley’s chorus section is the end of something
and thus does, indeed, act as a chorus in relation to the material that precedes it. A verse-
chorus reading is reinforced by the harmonic relationship between the two sections as well.
Specifically, the verse begins on tonic and moves away (to the dominant), whereas the chorus
beings away (on the submediant) and moves back to tonic (via overlap). Our sense of
beginning and ending (and thus of verse and chorus) is thereby affected by the large-scale
tonal motion itself.
All in all, the labels that Temperley chooses may be the most appropriate for this
song, especially considering all of the complicating factors that go into this decision. Yet one
should note how the focal quality in this song seems to exhibit a shift away from the second
Chapter 5: Blends 272
section. Perhaps the scale has not yet been tilted such that we perceive the first section to be
the more focal passage, but the stage appears to be set for the first section to convert to a
chorus. The nascent chorus-like quality of this first section may, in fact, explain the small
wrinkle in the succession pattern of the song. As mentioned before, the large-scale grouping
structure for “Tears in Heaven” can be seen as an AABAA form. The wrinkle, of course, is that
the A group appears not once but twice after the B group. Yet notice that the A group
immediately following the B group denies the listener of a true return of the first section (we
hear only an instrumental version). If we consider Temperley’s verse material to be vying for
(if not fully attaining) status as the focal musical material of the song, then the lack of a true
iteration of this section in the first A group after the vocal bridge arguably calls for the return
of a complete A group iteration. The fact that the second section returns as well at the end of
the song may simply derive from the interconnected nature of the two parts.
A conversion of blends
In “Tears in Heaven,” the refrain at the end of the second section was an important
factor in maintaining the sense of verse-chorus organization in the face of some evidence to
the contrary. The overlap of this refrain into the link material created a clear sense that the
section with the refrain should be considered as coming after its preceding section instead of
the other way around. (It is difficult to posit a supersection that would begin with the off-
tonic material, include the link, and then end with the more tonic-centered material.)
Moreover, the refrain itself acted as a focal feature and thereby imparted the off-tonic section
as a whole with a sense that it was the most important or most central passage of the song as a
whole. Yet given a very similar melodic-harmonic organization structure, our notion of form
may drastically shift if this type of refrain is no longer present.
Consider, for example, the song “That Feel” (Tom Waits, 1992). The song consists of
– for all intents and purposes – only two section types. The first section (let us call it the “A”
section) is shown in Example 5.5.04, and the second section (let us call it “B”) is shown in
Example 5.5.05. The entire succession pattern for the song is simply ABABABA (with a
closing fade on further A material); in other words, the two sections simply alternate back and
forth without any intermediary material.
Chapter 5: Blends 273
Example 5.5.04: “That Feel” (Tom Waits, 1992); first section (A)
Example 5.5.05: “That Feel” (Tom Waits, 1992); second section (B)
Comparing the construction of these two sections to the two main sections in “Tears
in Heaven,” we find remarkable similarities. As in the Eric Clapton song, the first section of
“That Feel” clearly prolongs tonic. Note as well that the harmonic structure in the first
section of both songs ends on a dominant chord. The off-tonic, unstable quality of the second
section of “That Feel” is also strong, despite the two instances of tonic harmony within these
eight bars. We find similarities between the lyric organizations of both songs as well. For
example, the lyrics to the off-tonic material in “That Feel” show no internal or external
pattern of repetition. But even more striking is how closely the lyric organization in the first
section of “That Feel” matches the first section of “Tears in Heaven.” Both repeat a title-
containing line at the end of each 4-bar hypermeasure, and this parallel structure continues
throughout the song. “That Feel” increases the level of repetition even further, though, as the
various iterations of this first section always repeat the lyrics in the first 4-bar hypermeasure
exactly (and some iterations consist only of repetitions of this first line).
The primary difference between the main material in “That Feel” and “Tears in
Heaven” is the lack of a refrain at the end of the off-tonic material. Consider the implications
that this difference has on our perception of section roles. In “That Feel,” it seems undeniable
chorus-like sections of stable harmonic material. These situations can be seen to reflect
bridge-like material acting in the role of a verse. Or, as in the case of “Today,” the situation
becomes highly complex, as there are a number of ways of parsing the song form.
Other songs present similar organizational schemes in which bridge-like material
seems to take on a verse-like role. Consider, for example, the song “Rehab” (Amy Winehouse,
2007). The main SRDC section of the song (first at 0:00) clearly stands as the chorus of the
song. In alternation with this SRDC material is a highly bridge-like passage (first at 0:27)
that serves as the sole transmission device for new lyric content. In “You’re One” (Imperial
Teen, 1996), similarly, the quiet tonic-initiated sections with repeating title-text lyrics (first at
0:29) alternate with loud off-tonic sections (first at 0:48) with new lyrics on each iteration. “I
Wanna Be Sedated” (Ramones, 1978), “I Don’t Wanna Grow Up” (Tom Waits, 1992), and “I
Feel the Earth Move” (Carole King, 1971) provide further instances in which our
understanding of section roles in the song potentially involves a blend of verse and bridge
qualities.
5.6: Summary
A central task of this chapter has been to show that section roles are not necessarily
mutually exclusive. There are many situations, in fact, where choosing a single section role to
describe a musical passage cannot fully capture the complex way this passage interacts with
our perception of section roles. Having to choose a single section role thus represents a type
of logical fallacy – the fallacy of false choice. Our understanding and perception of many
situations in rock songs can be adequately explained only by invoking multiple section roles.
This aspect of form derives from the fact that the mental processes we use to categorize
various section roles are always active. When we hear something that sounds very chorus-
like, for example, we do not turn off our ability to perceive prechorus-like qualities.
That being said, current theoretical systems of categorization for rock music
encourage us to seek out three primary roles: verse, chorus, and bridge. As a result, we may
often use these main role labels in cases where something more complex and nuanced is
going on. In particular, there are many instances where a musical passage evinces a blend of
two or more section roles. A number of different blends were discussed in this chapter, and
these blends were organized around the primary section roles of verse, chorus, and bridge
since at least one of these main section roles is involved in every blend.
In a general way, blends inherently involve ambiguity between section roles.
Ambiguity between section roles was, in fact, a central aspect of the previous chapter on
conversions. The discussion of conversions examined how one organizational scheme (such
Chapter 5: Blends 285
as a classic 16-bar SRDC) could expand (both literally and metaphorically) beyond a single
role to encompass multiple section roles. The discussion of blends could be considered as
something like the reverse of this process. In particular, many blends could be
conceptualized as a telescoping of song form. Take as a starting point Example 5.6.01a, for
example, in which a prototypical succession pattern for a few section roles is shown. In
Example 5.6.01b, this succession pattern has collapsed, such that the chorus and prechorus
roles have merged (e.g., “Communication Breakdown”). In a similar way, Example 5.6.01c
shows how the telescoping effect would cause the chorus and link roles to merge (e.g.,
“Jump”). We could also consider the case of “That’ll Be the Day” to be represented by
Example 5.6.01d, in which the verse and prechorus roles slip together. As can be seen, these
blends retain a standard sequence of section roles; it is simply that two neighboring sections
roles have been merged into one.
Example 5.6.01: Common blends involving neighboring section roles
Other types of blends involve our ability to perceive section roles on different levels of
grouping. This aspect was a central feature of blends that involved the bridge role. In such
situations, local evidence for a section role (e.g., verse, chorus, or prechorus) interacts with
global evidence for a bridge role. A robust understanding of the song structure thus requires
an ability to see multiple layers of form. The existence of multiple layers of section roles could
Verse Chorus Link
VersePrechorus
ChorusLink
VerseChorus
Link
VerseChorus
Prechorus
Prechorus
Prechorus
a)
b)
c)
d)
Chapter 5: Blends 286
be extended to verse blends as well, since the prechorus role may or may not be something
that acts within a larger verse context.
287
Chapter 6: Epilogue
Those readers that have successfully made their way through the preceding pages and
spent time carefully considering each and every example deserve a hearty congratulations.
Discussing such a large topic as form in rock music is not a simple endeavor, and it cannot
rely on only a few select examples. At this point, it is worth taking a step back in order to have
a broad look at what has been discussed herein. It would be impossible to adequately recap
every insight, as each topic has involved a number of details and factors. We can, however,
make some general observations and extend our findings to possibilities for future work. It is
also worth considering a few potential criticisms that may have cropped up in the process.
To restate what should be patently clear by now, this dissertation has employed a
prototype-based approach to talk about form in rock music. Evidence of prototype effects
(i.e. asymmetries in category membership) can be found in the discussions of form in the
work of prior authors (as discussed in the first chapter). But this dissertation confronts these
effects more directly than has ever been done before. In particular, this dissertation takes as
its first principle the notion that the form labels we use represent conceptual categories.
From this starting point, we realize that definitions very poorly account for the way we
understand and apply these labels. Of course, exactly how we represent these conceptual
categories is something that is complex, messy, and most likely unknowable. At minimum,
no single approach currently accounts for all of the phenomena found in categorization
studies. But a prototype-based approach – i.e., one that captures the attributes of those
members that are judged to be more central to the category – offers one of the most
successful methods to date.
As a result, Chapter 3 embarked on what may be seen as the first large-scale study of
prototypical instances of section roles in rock music. One important realization that emerges
from this discussion is that section roles are not as highly differentiated as they might initially
appear. One theory of conceptual categories holds that we tend to create and organize
categories such that they are maximally discriminable from each other (Rosch and Mervis
1975, 575-6). Indeed, the section role labels we use may represent our best attempts to
partition the types of situations we encounter in rock music into categories that contrast with
one another in the greatest possible way. Nevertheless, we find that the prototypical
instances of these categories share a great deal in common. For example, prototypical
prechorus and bridge sections both involve unstable harmonic content that lacks tonal
closure. Similarly, we find lyric repetition on future iterations to be a central attribute of both
the prechorus and chorus categories. Because a great deal of shared attributes can be found
Chapter 6: Epilogue 288
among prototypical section roles, there is a high potential for ambiguity with regard to the
role a particular passage plays within a song.
This high potential for ambiguity can also be found to derive from the fact that most
songs do not contain clear instances of every single type of section role. There thus exists a
continuum, not only in terms of where a particular passage lies with regard to its level of
evoking one section role or another, but also in terms of where a particular song lies with
regard to how well it evinces a set of section roles. This continuum between different song
forms was the focus of Chapter 4. As points of reference, this chapter used three common
organizational schemes: the 12-bar blues, the 16-bar SRDC, and the 32-bar AABA. From clear
instances of these schemes (in essence, prototypes), we could trace potential paths between
simple (or simpler) form types to more complex form types. The term “conversion” was
introduced in order to discuss how our perception of section roles was shifted in these
situations. On a higher level, though, we can think of conversion as a useful concept to
describe shifts in our perception of large-scale form types. We saw, for example, multiple
ways that a classic AABA form could be converted into a verse/chorus form.
One should note that the organizational schemes in Chapter 4 provide excellent
models for a class in popular music composition or songwriting, in that these schemes strike a
good balance between abstract and concrete melodic-harmonic elements. In this regard, it
would be nice to develop additional organizational schemes. One could imagine a host of
different organizational schemes, which taken together would represent some of the most
common ways of structuring the content of a rock song. We saw evidence of other potential
schemes in Chapter 5, in fact. The doubled tail refrain, for example, could be considered a
scheme (or at least a partial scheme) since it offers a similar mix of abstract and concrete
elements in the melodic and harmonic domains. Other blends discussed in Chapter 5, such as
the link/chorus, could also be thought of in terms of an organizational scheme.
Overall, the take-home message from Chapter 5 was that it is often counterproductive
to think of section roles as mutually exclusive labels. Indeed, the most appropriate way to
represent many recurring moments in rock music often seems to be via a blend of section
roles. A useful analytical method in this regard is the “multivalent” approach described by
James Webster (most recently in Bergé 2009). Yet it is uncommon, if not extremely rare, for
a music analyst to provide multiple form charts of a single rock song. Understandably, we
want to represent the form of a song in the simplest and most straightforward manner
possible. This is certainly a valuable desire in its own right. Doing so, however, we can
downplay what are often the most interesting and fascinating aspects of the song. One should
recognize that – given the standard set of section roles – songwriters are working within a
highly limited palette of form and section types. A central way to generate something unique
Chapter 6: Epilogue 289
or compelling, therefore, is to blend aspects of these form and section types. It is doubtful
that songwriters employ these blends in a conscious way, since these techniques have
received little attention in either scholarly or non-scholarly writing. At the same time,
though, it is doubtful that songwriters are not trying (at least sometimes) to present
something new and different to the listener. In this regard, blends provide a way to work
within the formal system while not falling into the trap of sounding overly derivative. A few
common types of blends were discussed in Chapter 5, but many more scenarios likely exist
within rock music.
Similarly, it is important to remind the reader that, although many prototypical
situations were described over the course of this dissertation, these situations are not
necessarily the only ones that account for our perception of a particular category. In the case
of the category “refrain,” for example, two related yet distinct subtypes were proposed: the tail
refrain and the head refrain. These subtypes were described in a number of domains,
including melodic phrase rhythm, harmony, lyrics, and placement within a larger grouping
structure. While it was posited that these are clear examples of the category “refrain,” there
may be other prototypical configurations that also trigger our sense of refrain quality in
similarly strong ways. Again, we should be careful not to lapse into thinking that any and all
refrain-like situations are equal. Recall, for instance, that the tail refrain was presented as a
more central member of the refrain category than the head refrain. So while other common
instantiations of categories may be found, we should weigh how they relate to other members
of that category.
In general, there is an underlying issue as to how we might more solidly know which
members and attributes are more central to our perception of any given category. Since this
dissertation did not involve an experimental component, many (if not all) of the attributes
and prototypical instances presented herein derive from communal if not personal
conjecture. It would be nice to test the insights from this dissertation within a controlled
setting, if only to confirm that they generally hold true. In this regard, the field of music
cognition holds great potential to reveal, validate, or disprove how particular features affect
our perception of categories in rock music. Given a verse-like passage in a minor mode, for
example, we might ask: to what extent does a move to an Ionian tonic engender chorus
quality? This is a complicated question, of course, as it involves a number of aspects. These
aspects include the frequency of the Ionian tonic, its placement within the hypermetric
context, and the total length of time spent on the Ionian tonic. The stimuli in such research
would necessarily require real music (as opposed to something like a probe tone), and so
controlling for a variety of confounding factors would be quite challenging. As one changes
the harmony in a single bar, for example, one affects a number of other parameters, such as
Chapter 6: Epilogue 290
the harmonic setting of the melody and the function of the surrounding chords within the
phrase as a whole. Nevertheless, there is a severe dearth of research in music cognition that
directly tackles perceptual issues within rock. It is thus worth embarking on this body of
work, if only because the field lays so wide open and so much remains to be discovered.
One advantage of conducting music cognition research on rock music is that the
general public has a relatively high exposure to and familiarity with rock (at least as opposed
to music of the common-practice era). As a result, studies would not need to be limited to the
relatively small subject pool of trained musicians often used in music cognition experiments.
In fact, the results from a cognitive study on rock music might be suspect if the study did not
utilize untrained musicians, since the majority of people who listen to (and possibly create)
rock music have little to no formal musical education. This advantage comes with its own set
of problems, though. An untrained musician, for example, does not necessarily have the same
ability to explicitly describe his or her response to music as does a trained musician, who has
access to a large technical vocabulary. Yet the technical vocabulary of rock music is not
something with which even a trained musician has extensive experience, since rock music is
only briefly discussed (if at all) within a standard music education. The ideal subject for a
study of form in rock music might thus be the practicing rock musician. It is somewhat
difficult to measure what constitutes a practicing rock musician, however (as opposed to
simply measuring years of formal music education), and so even this solution carries its own
set of issues.
Matters of subject pool aside, music cognition holds the potential to answer a number
of questions left unresolved in this dissertation. One central topic was how much music
should be considered to constitute a single measure. To some readers, this issue may seem
somewhat trivial. But one should recall how useful melodic phrase rhythms proved to be in
Chapter 4 with regard to tracking conversions of various organizational schemes. These
organizational schemes were characterized in strong part by the particular way that the vocal
melody was grouped within the prevailing hypermetric structure. Measure length
calculations thus impact our ability to reference a given song back to a particular
organizational scheme. Recent work on phrase rhythm in jazz music (Love 2011) further
highlights the importance of the relationship between melodic grouping structures and the
hypermetric framework in popular music. Some enhanced guidelines for the music analyst
would be useful in this regard, since the current system of assigning measure lengths (via the
snare drum pattern or backbeat) is not always reliable (as discussed in Chapter 3). Previous
research on the perception of absolute time (as summarized in London 2004, 27ff) offers
some initial data in relation to measure lengths. In particular, studies as to where listeners
feel “the beat” relate to questions of measure lengths. Nonetheless, measures do not
Chapter 6: Epilogue 291
necessarily contain an equivalent number of beats, and so it is not clear how to directly apply
data on pulse salience, for example, to measure lengths. Our sense of a measure inherently
involves more factors than simply the beat alone, and these factors probably include the
harmonic rhythm, the level of rhythmic subdivision in the melody, the melodic phrase
lengths, and the overall tonal framework. There is a good chance, moreover, that these
factors interact in ways that make it somewhat difficult to isolate one from the other. At the
end of the day, it is doubtful that we can come up with a system in which ambiguous cases will
cease to exist. Yet further study on this issue is certainly warranted, since the success of
future work on melodic phrase rhythm in rock music is heavily reliant on a consistent metric
for measure lengths.
Research involving melodic phrase rhythm in rock music includes its own set of
problems as well. One fundamental concern is where melodic phrase groupings begin and
end. As opposed to purely instrumental music, the vast majority of rock music has the
advantage of lyrics to help in the identification of melodic phrase beginnings and endings.
Yet we cannot assume that groupings in the lyrics and melody are always aligned. Breaks (or
rests) in the melody seem to be the most obvious way of parsing the grouping structure, yet
we have seen many cases that thwart any simple mechanism based on this parameter alone.
Sometimes, a new melodic phrase begins immediately after the end of the prior vocal phrase.
In other instances, it is useful to posit a single overarching vocal phrase despite a number of
internal pauses and interruptions. Of course, the way in which a melodic phrase is grouped
with regard to the underlying hypermeter is, fundamentally speaking, a matter of analytical
preference. In fact, one aspect of this dissertation with which readers might take issue
concerns the particular melodic groupings that have been presented herein.
Some readers may raise objections with other aspects of this work as well. For
example, one might feel that an undue emphasis has been placed on the importance and value
of the section labels themselves. In the analyses within this dissertation, much effort was
exerted to appreciate what label (or labels) might be most appropriate for a particular span of
music. To some, this effort may seem somewhat tangential to the higher-level act of music
analysis itself. For example, Stephenson writes, “It must be remembered that the point of
formal analysis… is not to identify passages with the proper labels, for the labels serve merely
as points of perspective around which to organize our thoughts” (2002, 133). Stephenson
goes on to say that it does not matter whether we call some particular passage, for instance, a
short chorus or a long refrain. What matters is that we have drawn the reader’s attention to a
specific feature of a specific moment in the music, and that this feature serves as a starting
point for further discussion.
Chapter 6: Epilogue 292
Indeed, it would be hard to argue that labeling the various parts of a song is an
endpoint in the analytical process (any more than a full-fledged Schenkerian graph is the
endpoint in the analysis of a Beethoven sonata). Yet a single section label undoubtedly
conveys a significant amount of meaning, and this meaning undeniably influences (at least to
a certain extent) our view of the song. For instance, a central principle that drives many
analyses is the explanation of interesting or unique moments within a piece of music. It is
difficult to apprehend what is interesting and unique, though, without a solid understanding
of what is prototypical and commonplace. Stephenson is certainly correct that the
distinction, for instance, between whether a particular passage is a short chorus or a long
refrain can often seem trivial. But this is exactly what it interesting about these types of
situations. Calling the passage one or the other draws our attention away from the fact that
such situations represent an inextricable blend of chorus and refrain roles. To be fair,
Stephenson is not necessarily arguing against a “both/and” reading; he seems merely to be
saying that labels have only limited analytical value. Undoubtedly, this is true. At the same
time, however, we should not undervalue the labels we do have, for they contain powerful
associations and implications.
In a related manner, some readers may feel that this dissertation does not reflect the
personal, subjective, and contextual interpretations of section labels. Since each person
inherently interacts with and listens to a different body of music during their lifetime, one
could argue that we have each developed highly personalized notions of what constitutes good
and bad examples. Consequently, no single shared understanding of prototypical situations
can be said to exist. Even when two people are familiar with the same song, we cannot
assume that they have had the same experience with that song or are influenced by that song
in the same way. Someone who is an avid Beatles fan, for example, may have an entirely
different understanding of what constitutes a prototypical verse as compared to someone who
listens primarily to heavy metal. This dissertation, it could be argued, represents only the
personal, subjective, and contextual interpretations of its author. One might say, in other
words, that the work herein does not model how we hear, but rather how just a single person
hears.
In this regard, it is worth reminding the reader that, while two individuals may
disagree as to the exact boundaries of a category, it is widely confirmed that people
overwhelmingly agree on judgments of those more central members (see Ch. 2; Rosch 1978).
The judgments of prototypes made in Ch. 3 should thus be considered to reasonably reflect
the judgments of listeners in general. Nevertheless, this potential criticism raises a valid
point. Different eras and styles of rock music undoubtedly utilize different strategies for
organizing musical material. Chorus sections in 1950s blues music are certainly of a different
Chapter 6: Epilogue 293
type than those typically found in the music of Nirvana. Perhaps it is not fair to judge one
style in the same context as the other. In using the term “chorus” to describe both passages,
however, we show evidence that we do hear one in the context of the other – that these
musical passages are related in some way. This is not to say that what triggers our sense of
chorus quality is necessarily the same in both songs. Rather, our understanding of a section
role in one style overlaps and influences our understanding of that role in another. For this
reason, style analysis has been and will continue to be a valuable and worthwhile endeavor.
Some efforts have been made in this dissertation along these lines (such as the distinction
between a classic bridge and a modern bridge), and useful insights were gained as a result
(such as the similarity between a classic bridge and a prechorus with regard to the conversion
of AABA forms).
In conclusion, this dissertation is not meant to stand as the final word on prototypical
configurations in rock music. This dissertation hopes instead to refocus the discourse on
form in rock music via a prototype-based approach. As stated in the opening paragraph of
this study, “form” has been one of the most persistently complex topics facing theorists of
classic music. Form in rock music, one should agree, is no different.
294
Bibliography
n.a. 2004. “The 500 greatest songs of all time.” Rolling Stone 963 (Dec. 9): 65-165. n.a. n.d. “Blitzkrieg Bop.” Wikipedia. http://en.wikipedia.org/wiki/Blitzkrieg_Bop (accessed
August 22, 2011). n.a. n.d. “Today (The Smashing Pumpkins Song).” Wikipedia.
http://en.wikipedia.org/wiki/Today_(The_Smashing_Pumpkins_song) (accessed August 22, 2011).
n.a. n.d. “Train in Vain.” Wikipedia. http://en.wikipedia.org/wiki/Train_in_Vain (accessed
August 22, 2011). Agmon, Eytan. 1995. “Functional Harmony Revisited: A Prototype-Theoretic Approach.”
Music Theory Spectrum 17/2 (Autumn): 196-214. Bergé, Pieter, ed. 2009. Musical Form, Forms, and Formenlehre: Three Methodological
Reflections. Leuven, Belgium: Leuven University Press. Berlin, Brent and Paul Kay. 1969. Basic Color Terms: Their Universality and Evolution.
Berkeley, CA: University of California Press. Biamonte, Nicole. 2010. “Triadic Modal and Pentatonic Patterns in Rock Music.” Music
Theory Spectrum. 32/2 (Fall): 95-110. Bonds, Mark. 1991. Wordless Rhetoric: Musical Form and the Metaphor of the Oration.
Cambridge, MA: Harvard University Press. Brown, Matthew. 2004. “Composing with Prototypes: Charting Debussy’s L’Isle joyeuse.”
Intégral 18-19: 151-88. . 2005. Explaining Tonality: Schenkerian Theory and Beyond. Rochester, NY:
University of Rochester Press. Brown, Matthew, Douglas Dempster, and Dave Headlam. 1997. “The #IV(bV) Hypothesis:
Testing the Limits of Schenker’s Theory of Tonality.” Music Theory Spectrum 19/2 (Autumn): 155-83.
Burnham, Scott. 2002. “Form.” The Cambridge History of Western Music Theory. Edited by
Thomas Christensen. Cambridge: Cambridge University Press, 880-906. Burns, Lori. 2005. “Meaning in a Popular Song: The Representation of Masochistic Desire in
Sarah McLachlan’s ‘Ice.’” In Stein 2005, 136-48. . 2008. “Analytic Methodologies for Rock Music: Harmonic and Voice-Leading
Strategies in Tori Amos’s ‘Crucify.’” In Everett 2008a, 63-92. Caplin, William. 1998. Classical Form: A Theory of Formal Functions for the Instrumental
Music of Haydn, Mozart, and Beethoven. Oxford: Oxford University Press.
Bibliography 295
Cohn, Richard. “Introduction to Neo-Riemannian Theory: A Survey and a Historical Perspective.” Journal of Music Theory 42/2 (Autumn): 167-80.
Cook, Nicholas. 1987. A Guide to Musical Analysis. New York: W. W. Norton. Covach, John. 2003. “Pangs of history in late 1970s new-wave rock.” In Analyzing Popular
Music. Edited by Allan F. Moore. Cambridge: Cambridge University Press. . 2005. “Form in Rock Music: A Primer.” In Stein 2005, 65-76. . 2006. “From ‘Craft’ to ‘Art’: Formal Structure in the Music of the Beatles.” In
Reading the Beatles: Cultural Studies, Literary Criticism, and the Fab Four. Edited by Kenneth Womack and Todd F. Davis. Albany: State University of New York Press, 37-54.
. 2009. What’s That Sound?: An Introduction to Rock and Its History, 2nd ed. New
York: W. W. Norton. . 2010. “Lieber and Stoller, the Coasters, and the ‘Dramatic AABA’ Form.” In
Sounding Out Pop: Analytical Essays in Popular Music. Edited by Mark Spicer and John Covach. Ann Arbor, MI: University of Michigan Press, 1-17.
Covach, John and Graeme Boone, eds. 1997. Understanding Rock: Essays in Musical
Analysis. Oxford: Oxford University Press. Davis, Sheila. 1985. The Craft of Lyric Writing. Cincinnati, OH: Writer’s Digest Books. de Clercq, Trevor and David Temperley. 2011. “A corpus analysis of rock harmony.” Popular
Music 30/1 (January): 47-70. Denham, Kristin and Anne Lobeck. 2009. Linguistics for Everyone: An Introduction. Boston,
MA: Wadsworth. Doll, Christopher. 2007. “Listening to Rock Harmony.” Ph.D. dissertation, Columbia
University. . 2011. “Rockin’ Out: Expressive Modulation in Verse-Chorus Form.” Music Theory
Online 17/3 (October). Endrinal, Christopher. 2008. “Form and Style in the Music of U2.” Ph.D. dissertation, Florida
State University. Everett, Walter. 1999. The Beatles as Musicians: Revolver through the Anthology. Oxford:
Oxford University Press. . 2001. The Beatles as Musicians: The Quarry Men through Rubber Soul. Oxford:
Oxford University Press. . 2004. “Making Sense of Rock’s Tonal Systems.” Music Theory Online. 10/4
(December). , ed. 2008a. Expression in Pop-Rock Music: Critical and Analytical Essays, 2nd ed.
New York: Routledge.
Bibliography 296
. 2008b. “Pitch Down the Middle.” In Everett 2008a, 111-74. . 2009. The Foundations of Rock: From “Blue Suede Shoes” to “Suite: Judy Blue
Eyes.” Oxford: Oxford University Press. Gotlieb, Heidi and Vladimir Konečni. 1985. “The effects of instrumentation, playing style, and
structure in the Goldberg Variations by Johann Sebastian Bach.” Music Perception 3/1 (Fall): 87-102.
Gjerdingen, Robert. 2007. Music in the Galant Style. Oxford: Oxford University Press. Harnad, Stevan, ed. 1987. Categorical perception: the groundwork of cognition. Cambridge:
Cambridge University Press. Harris, K. Paul. 2006. “U2’s Creative Process: Sketching in Sound.” Ph.D. dissertation,
University of North Carolina, Chapel Hill. Harrison, Daniel. 1997. “After Sundown: The Beach Boys’ Experimental Music.” In Covach
and Boone 1997, 33-57. Hepokoski, James and Warren Darcy. 2006. Elements of Sonata Theory: Norms, Types, and
Deformations in the Late-Eighteenth-Century Sonata. Oxford: Oxford University Press.
Hubbs, Nadine. 2008. “The Imagination of Pop-Rock Criticism.” In Everett 2008a, 215-37. Hurley, Patrick J. 2008. A Concise Introduction to Logic, 10th ed. Belmont, CA: Thomson
Wadsworth. Karno, Mitchell and Vladimir Konečni. 1992. “The Effects of Structural Interventions in the
First Movement of Mozart’s Symphony in G Minor K. 550 on Aesthetic Preference.” Music Perception 10/1 (Fall): 63-72.
Konečni, Vladimir. 1984. “Elusive effects of artists’ ‘messages.’” In Cognitive processes in the
perception of art. Edited by W. R. Crozier and A. J. Chapman. Amsterdam: Elsevier Science Publishers, 71-96.
Koozin, Timothy. 2008. “Fumbling Towards Ecstasy: Voice Leading, Tonal Structure, and the
Theme of Self-Realization in the Music of Sarah McLachlan.” In Everett 2008a. Lakoff, George. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about
the Mind. Chicago: The University of Chicago Press. Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. Chicago: University of
Chicago Press. Lerdahl, Fred and Ray Jackendoff. 1983. A Generative Theory of Tonal Music. Cambridge,
MA: The MIT Press. London, Justin. 2004. Hearing in Time: Psychological Aspects of Musical Meter. Oxford:
Oxford University Press.
Bibliography 297
Love, Stefan. 2011. “On Phrase Rhythm in Jazz.” Ph.D. dissertation, Eastman School of Music.
Martin, George with Jeremy Hornsby. 1979. All You Need Is Ears. New York: St. Martin’s
Press. McClary, Susan and Robert Walser. 1990. “Start Making Sense! Musicology Wrestles with
Rock.” In On Record: Rock, Pop, and the Written Word. Edited by Simon Frith and Andrew Goodwin. London: Routledge, 277-92.
McDonald, Chris. 2000. “Exploring modal subversions in alternative music.” Popular Music
19/3 (October): 355-63. Medin, Douglas and Marguerite Schaffer. 1978. “Context theory of classification learning.”
Psychological Review 85/3 (May): 207-38. Middleton, Richard. 1990. Studying Popular Music. Buckingham: Open University Press. Moore, Allan. 1992. “Patterns of Harmony.” Popular Music 11/1 (January): 73-106. . 2001. Rock: The Primary Text: Developing a musicology of rock, 2nd ed.
Aldershot, UK: Ashgate. Murphy, Gregory. 2002. The Big Book of Concepts. Cambridge, MA: MIT Press. Neal, Jocelyn. 2007. “Narrative Paradigms, Musical Signifiers, and Form as Function in
Country Music.” Music Theory Spectrum 29/1 (Spring): 41-72. . 2008. “Country-Pop Formulae and Craft: Shania Twain’s Crossover Appeal.” In
Everett 2008a, 285-311. Osborn, Bradley. 2010. “Beyond Verse and Chorus: Experimental Formal Structures in Post-
Millennial Rock Music.” Ph.D. dissertation, University of Washington. Parncutt, Richard. 1994. “A Perceptual Model of Pulse Salience and Metrical Accent in
Musical Rhythms.” Music Perception 11/4 (Summer): 409-64. Pattison, Pat. 1991. Songwriting: Essential Guide to Lyric Form and Structure: Tools and
Techniques for Writing Better Lyrics. Boston: Berklee Press. Perricone, Jack. 2000. Melody in Songwriting: Tools and Techniques for Writing Hit Songs.
http://www.recmusicbeatles.com/public/files/awp/awp.html (accessed May 17, 2011).
Rips, Lance. 1975. “Inductive judgments about natural categories.” Journal of Verbal
Learning and Learning Behavior 14/6 (December): 665-81. Rips, Lance, Edward Shoben, and Edward Smith. 1973 “Semantic distance and the
verification of semantic relations.” Journal of Verbal Learning and Learning Behavior 12/1 (February): 1-20.
Bibliography 298
Rosch, Eleanor. 1973. “Natural Categories.” Cognitive Psychology 4: 328-50. . 1975a. “Cognitive Representations of Semantic Categories.” Journal of
Experimental Psychology: General 104/3: 192-233. . 1975b. “Cognitive Reference Points.” Cognitive Psychology 7: 532-47. . 1978. “Principles of Categorization.” In Cognition and Categorization. Edited by
Eleanor Rosch and Barbara Lloyd. Hillsdale, NJ: Lawrence Erlbaum Associates, 27-48.
Rosch, Eleanor and Carolyn Mervis. 1975. “Family Resemblances: Studies in the Internal
Structure of Categories.” Cognitive Psychology 7: 573-605. Rothstein, William. [1989] 2007. Phrase Rhythm in Tonal Music, reprint edition. Ann Arbor,
MI: Musicalia Press. Schachter, Carl. [1987] 1999. “Aspects of Meter.” In Unfoldings: Essays in Schenkerian
Theory and Analysis. Oxford: Oxford University Press, 79-117. Schenker, Heinrich. [1935] 1979. Free Composition. Translated and edited by Ernst Oster.
Hillsdale, NY: Pendragon Press. Smith, Edward and Douglas Medin. 1981. Categories and Concepts. Cambridge, MA: Harvard
University Press. Smith, J. David and John Minda. 1998. “Prototypes in the Mist: The Early Epochs of Category
Learning.” Journal of Experimental Psychology 24/6: 1411-36. Spicer, Mark. 2005. Review of Everett 2001. Music Theory Online 11/4 (October). Stein, Deborah, ed. 2005. Engaging Music: Essays in Music Analysis. Oxford: Oxford
University Press. Stephan-Robinson, Anna. 2009. “Form in Paul Simon’s Music.” Ph.D. dissertation, Eastman
School of Music. Stephenson, Ken. 2002. What to Listen for in Rock: A Stylistic Analysis. New Haven: Yale
University Press. Summach, Jay. 2011. “The Structure, Function, and Genesis of the Prechorus.” Music Theory
Online 17/3 (October). Tagg, Philip. 1982. “Analysing Popular Music: Theory, Method and Practice.” Popular Music
2: 37-67. Temperley, David. 1999. “Syncopation in rock: a perceptual perspective.” Popular Music 18/1:
19-40. . 2001. The Cognition of Basic Musical Structures. Cambridge, MA: The MIT Press. . 2007. “The Melodic-Harmonic ‘Divorce’ in Rock.” Popular Music 26/2: 323-42.
Bibliography 299
. 2010. Online analyses of 200 rock songs. http://theory.esm.rochester.edu/rock_corpus/dt_analyses/ (accessed January 31, 2011).
. 2011. “The Cadential IV in Rock.” Music Theory Online 17/1 (March). Tillmann, Barbara and Emmanuel Bigand. 1996. “Does Formal Musical Structure Affect
Perception of Musical Expressiveness?” Psychology of Music 24/1 (April): 3-17. Whittall, Arnold. 2010. “Form.” In Grove Music Online. Oxford Music Online.
http://www.oxfordmusiconline.com/subscriber/article/grove/music/09981 (accessed March 22, 2010).
Winkler, Peter. 1997. “Writing Ghost Notes: The Poetics and Politics of Transcription.” In
Keeping Score: Music, Disciplinarity, Culture. Edited by David Schwarz, Anahid Kassabian, and Lawrence Siegel. Charlottesville, VA: University Press of Virginia, 169-203.
Wittgenstein, Ludwig. [1953] 2001. Philosophical Investigations. Translated by G. E. M.
Anscombe. Oxford: Blackwell. Zadeh, Lotfi A. 1965. “Fuzzy Sets.” Information and Control 8: 338-353. Zbikowski, Lawrence. 2002. Conceptualizing Music: Cognitive Structure, Theory, and
Analysis. Oxford: Oxford University Press.
300
Songs Cited
1951 “Hey Good Lookin’” Hank Williams 1951 “Violent Love” The Big Three 1952 “Honky Tonk Blues” Hank Williams 1953 “Your Cheatin’ Heart” Hank Williams 1953 “Money Honey” The Drifters 1954 “Shake, Rattle and Roll” Big Joe Turner 1954 “Evil” Howlin’ Wolf 1954 “When the Lights Go Out” Jimmy Witherspoon 1954 “Earth Angel” The Penguins 1955 “Maybellene” Chuck Berry 1955 “Tutti Frutti” Little Richard 1955 “Do Me Right” Lowell Fulson 1955 “Young Fashioned Ways” Muddy Waters 1955 “Ko Ko Mo (I Love You So)” The Crew Cuts 1955 “The Great Pretender” The Platters 1955 “Crazy for My Baby” Willie Dixon 1955 “Pain in My Heart” Willie Dixon 1956 “Roll Over Beethoven” Chuck Berry 1956 “Hound Dog” Elvis Presley 1956 “Love Me” Elvis Presley 1956 “Blue Suede Shoes” Elvis Presley 1956 “Tryin’ to Get to You” Elvis Presley 1956 “Long Tall Sally” Little Richard 1956 “Ooby Dooby” Roy Orbison 1956 “Turtle Dovin’” The Coasters 1956 “In the Still of the Night” The Five Satins 1956 “One in a Million” The Platters 1957 “Butterfly” Andy Williams 1957 “Everyday” Buddy Holly 1957 “Peggy Sue” Buddy Holly 1957 “Jailhouse Rock” Elvis Presley 1957 “I’m Walkin’” Fats Domino 1957 “Great Balls of Fire” Jerry Lee Lewis 1957 “You Don’t Owe Me a Thing” Johnnie Ray 1957 “You Send Me” Sam Cooke 1957 “Young Blood” The Coasters 1957 “That’ll Be the Day” The Crickets 1957 “Come Go with Me” The Del-Vikings 1958 “Rockin’ Robin” Bobby Day 1958 “Johnny B. Goode” Chuck Berry 1958 “Donna” Ritche Valens 1958 “All I Have to Do Is Dream” The Everly Brothers 1958 “Born Too Late” The Poni-Tails 1959 “Dream Lover” Bobby Darin 1959 “What'd I Say” Ray Charles 1959 “Charlie Brown” The Coasters 1959 “That Is Rock and Roll” The Coasters 1960 “Devil or Angel” Bobby Vee 1960 “True Love Ways” Buddy Holly 1960 “Stuck on You” Elvis Presley
Songs Cited 301
1960 “Handy Man” Jimmy Jones 1960 “Shop Around” The Miracles 1960 “Boys” The Shirelles 1960 “Will You Love Me Tomorrow” The Shirelles 1961 “Stand by Me” Ben E. King 1961 “Tossin’ and Turnin’” Bobby Lewis 1961 “Crazy” Patsy Cline 1961 “Blue Moon” The Marcels 1961 “Mama Said” The Shirelles 1963 “Dead Presidents” Little Walter 1963 “Little Deuce Coupe” The Beach Boys 1963 “Surfer Girl” The Beach Boys 1963 “Surfin’ USA” The Beach Boys 1963 “From Me to You” The Beatles 1963 “I Want to Hold Your Hand” The Beatles 1963 “She Loves You” The Beatles 1963 “Chains” The Beatles 1963 “I Saw Her Standing There” The Beatles 1963 “Love Me Do” The Beatles 1963 “P. S. I Love You” The Beatles 1963 “Please Please Me” The Beatles 1964 “A Change is Gonna Come” Sam Cooke 1964 “I Feel Fine” The Beatles 1964 “A Hard Day’s Night” The Beatles 1964 “And I Love Her” The Beatles 1964 “Can’t Buy Me Love” The Beatles 1964 “I’ll Cry Instead” The Beatles 1964 “You Can’t Do That” The Beatles 1964 “Eight Days a Week” The Beatles 1964 “Everybody’s Trying to Be My Baby” The Beatles 1964 “I’ll Follow the Sun” The Beatles 1964 “You’ve Lost That Lovin’ Feelin’” The Righteous Brothers 1964 “(The Best Part of) Breakin’ Up” The Ronettes 1964 “Walking in the Rain” The Ronettes 1964 “Where Did Our Love Go” The Supremes 1965 “I’ve Got a Tiger by the Tail” Buck Owens 1965 “California Girls” The Beach Boys 1965 “Help!” The Beatles 1965 “Ticket to Ride” The Beatles 1965 “Yesterday” The Beatles 1965 “Drive My Car” The Beatles 1965 “As Tears Go By” The Rolling Stones 1965 “(I Can’t Get No) Satisfaction” The Rolling Stones 1966 “Jimmy Mack” Martha Reeves and the Vandellas 1966 “When a Man Loves a Woman” Percy Sledge 1966 “Good Vibrations” The Beach Boys 1966 “Taxman” The Beatles 1966 “I’m a Believer” The Monkees 1967 “Strange Brew” Cream 1967 “Sunshine of Your Love” Cream 1967 “Penny Lane” The Beatles 1967 “A Day in the Life” The Beatles 1967 “Lucy in the Sky with Diamonds” The Beatles
Songs Cited 302
1967 “So You Want to Be a Rock ‘n’ Roll Star” The Byrds 1967 “Light My Fire” The Doors 1967 “I Can See for Miles” The Who 1968 “Crossroads” Cream 1968 “I Heard It Through the Grapevine” Marvin Gaye 1968 “At the Zoo” Simon and Garfunkel 1968 “Hey Jude” The Beatles 1968 “La-La (Means I Love You)” The Delfonics 1968 “Jumpin’ Jack Flash” The Rolling Stones 1968 “Sympathy for the Devil” The Rolling Stones 1969 “Suspicious Minds” Elvis Presley 1969 “Communication Breakdown” Led Zeppelin 1969 “Whole Lotta Love” Led Zeppelin 1969 “Thank You (Falettinme Be Mice Elf Agin)” Sly and The Family Stone 1969 “Sin City” The Flying Burrito Brothers 1969 “Honky Tonk Women” The Rolling Stones 1970 “You’ve Got a Friend” Carole King 1970 “Travelin’ Band” Creedence Clearwater Revival 1970 “Layla” Derek and the Dominos 1970 “Get Up (I Feel Like Being a) Sex Machine” James Brown 1970 “Tangerine” Led Zeppelin 1971 “I Feel the Earth Move” Carole King 1971 “Stairway to Heaven” Led Zeppelin 1971 “What’s Going On” Marvin Gaye 1971 “Won’t Get Fooled Again” The Who 1972 “Smoke on the Water” Deep Purple 1972 “Take It Easy” Eagles 1973 “Bad, Bad Leroy Brown” Jim Croce 1974 “Take Me to the River” Al Green 1974 “September Gurls” Big Star 1974 “Haven’t Got Time for the Pain” Carly Simon 1974 “Let It Grow” Eric Clapton 1975 “Walk This Way” Aerosmith 1975 “Born to Run” Bruce Springsteen 1976 “More Than a Feeling” Boston 1976 “Hotel California” Eagles 1976 “Blitzkrieg Bop” Ramones 1977 “Feels Like the First Time” Foreigner 1977 “Handy Man” James Taylor 1977 “We Will Rock You” Queen 1977 “God Save the Queen” The Sex Pistols 1978 “Old Time Rock and Roll” Bob Seger & The Silver Bullet Band 1978 “I Wanna Be Sedated” Ramones 1978 “Just What I Needed” The Cars 1979 “Highway to Hell” AC/DC 1979 “Gotta Serve Somebody” Bob Dylan 1979 “Hot Stuff” Donna Summer 1979 “Train in Vain” The Clash 1980 “Him” Rupert Holmes 1980 “Don’t Stand So Close to Me” The Police 1981 “Hard to Say” Dan Fogelberg 1981 “Don’t Stop Believin’” Journey 1982 “Hungry Like the Wolf” Duran Duran
Songs Cited 303
1982 “Rio” Duran Duran 1982 “1999” Prince 1982 “Little Red Corvette” Prince 1982 “Should I Stay or Should I Go?” The Clash 1983 “Pink Houses” John Cougar Mellencamp 1983 “Billie Jean” Michael Jackson 1983 “Every Breath You Take” The Police 1983 “Wrapped Around Your Finger” The Police 1984 “Cover Me” Bruce Springsteen 1984 “Run to You” Bryan Adams 1984 “Summer of ‘69” Bryan Adams 1984 “Material Girl” Madonna 1984 “Cool It Now” New Edition 1984 “Magic” The Cars 1984 “You Might Think” The Cars 1984 “Jump” Van Halen 1985 “Sara” Starship 1986 “Livin’ on a Prayer” Bon Jovi 1986 “You Give Love a Bad Name” Bon Jovi 1986 “Papa Don’t Preach” Madonna 1987 “Angel” Aerosmith 1987 “Faith” George Michael 1987 “Sweet Child o’ Mine” Guns N’ Roses 1987 “Smooth Criminal” Michael Jackson 1987 “I Still Haven’t Found What I’m Looking For” U2 1988 “One” Metallica 1988 “Nothin’ But a Good Time” Poison 1989 “Rockin’ in the Free World” Neil Young 1990 “Screen Door” Uncle Tupelo 1991 “In Bloom” Nirvana 1991 “Smells Like Teen Spirit” Nirvana 1991 “The Walk” Sawyer Brown 1991 “Who’s Gonna Ride Your Wild Horses” U2 1992 “End of the Road” Boyz II Men 1992 “Nuthin’ But a ‘G’ Thang” Dr. Dre 1992 “Tears in Heaven” Eric Clapton 1992 “I Don’t Wanna Grow Up” Tom Waits 1992 “That Feel” Tom Waits 1992 “Crucify” Tori Amos 1993 “Mystery Dance” Elvis Costello 1993 “All Apologies” Nirvana 1993 “Today” The Smashing Pumpkins 1994 “She” Green Day 1994 “When I Come Around” Green Day 1994 “Elsewhere” Sarah McLachlan 1994 “All I Wanna Do” Sheryl Crow 1994 “Don’t Take the Girl” Tim McGraw 1994 “Buddy Holly” Weezer 1995 “1979” The Smashing Pumpkins 1996 “You’re One” Imperial Teen 1997 “Horses” Palace Music 1997 “Building a Mystery” Sarah McLachlan 1997 “Black Eyes, Blue Tears” Shania Twain
Songs Cited 304
1997 “Honey, I’m Home” Shania Twain 2000 “Beautiful Day” U2 2000 “Elevation” U2 2002 “In My Car (I’ll Be the Driver)” Shania Twain 2002 “Waiter! Bring Me Water!” Shania Twain 2003 “Why Can’t I” Liz Phair 2004 “Don’t Tell Me” Avril Lavigne 2004 “My Happy Ending” Avril Lavigne 2004 “Horses” Bonnie “Prince” Billy 2004 “Behind These Hazel Eyes” Kelly Clarkson 2004 “City of Blinding Lights” U2 2004 “Original of the Species” U2 2005 “Who We Are” Hope Partlow 2006 “Hero/Heroine” Boys Like Girls 2006 “The Great Escape” Boys Like Girls 2006 “Thunder” Boys Like Girls 2007 “Rehab” Amy Winehouse 2008 “Hot n Cold” Katy Perry 2008 “Love Story” Taylor Swift 2008 “You’re Not Sorry” Taylor Swift 2009 “Bad Romance” Lady Gaga 2009 “Where the Lines Overlap” Paramore 2009 “Naturally” Selena Gomez & the Scene 2009 “You Belong with Me” Taylor Swift 2010 “Just the Way You Are” Bruno Mars